UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Infection prevention and control and antibiotic stewardshipTwo sides of the same coin in the prevention of antimicrobial resistanceMaina, J.M.

Publication date2021Document VersionFinal published versionLicenseOther

Link to publication

Citation for published version (APA):Maina, J. M. (2021). Infection prevention and control and antibiotic stewardship: Two sides ofthe same coin in the prevention of antimicrobial resistance.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an opencontent license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, pleaselet the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the materialinaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letterto: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. Youwill be contacted as soon as possible.

Download date:20 Oct 2022

1

Infection prevention and

control and antibiotic

stewardship: two sides of

the same coin in the

prevention of antimicrobial

resistance

Jackson Michuki Maina

2

This thesis was prepared at the Faculty of Medicine, Academic Medical Center

of the University of Amsterdam, The Netherlands and at the KEMRI-Wellcome

Trust Research Programme, Nairobi Kenya.

Copyright 2021, J. M Maina, Nairobi, Kenya

All rights are reserved. No part of this thesis may be reproduced, stored or

transmitted in any form or by any means without the prior permission of the

author.

Front cover: Daniel Maina

Paranymph: Nduku Kiko

Printed by IPSKAMP with support from the Academic Medical Center, University

of Amsterdam.

Michuki was the recipient of an IDEAL (Initiative to Develop African research

Leaders) PhD fellowship. This work was supported through the DELTAS Africa

Initiative [DEL-15-003]. The DELTAS Africa Initiative is an independent funding

scheme of the African Academy of Sciences (AAS)'s Alliance for Accelerating

Excellence in Science in Africa (AESA) and supported by the New Partnership for

Africa's Development Planning and Coordinating Agency (NEPAD Agency) with

funding from the Wellcome Trust [107769/Z/10/Z] and the UK government. The

views expressed in this publication are those of the author(s) and not

necessarily those of AAS, NEPAD Agency, Wellcome Trust or the UK

government.

3

Infection prevention and control and antibiotic stewardship:

two sides of the same coin in the prevention of antimicrobial

resistance

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctor

aan de Universiteit van Amsterdam

op gezag van de Rector Magnificus

prof. dr. ir. K.I.J. Maex

ten overstaan van een door het College voor Promoties ingestelde commissie,

in het openbaar te verdedigen in de Agnietenkapel

op woensdag 24 februari 2021, te 16.00 uur

door Jackson Michuki Maina

geboren te Nyeri

4

Promotiecommissie

Promotores: prof. dr. C. Schultsz

prof. dr. M. English

AMC-UvA

University of Oxford

Copromotores: dr. O. Tosas Auguet

dr. J. McKnight

University of Oxford

University of Oxford

Overige leden: prof. dr. M. van Vugt

prof. dr. S.E. Geerlings

AMC-UvA

AMC-UvA

prof. dr. M. Boele van Hensbroek AMC-UvA

prof. dr. R.C. Pool Universiteit van Amsterdam

prof. dr. F.N. Were University of Nairobi

prof. dr. A. Voss Radboud Universiteit

Faculteit der Geneeskunde

5

TABLE OF CONTENTS

Abbreviations 6

Chapter 1 Introduction 7

Chapter 2 Extending the use of the World Health Organisations’

water sanitation and hygiene assessment tool for surveys

in hospitals – from WASH-FIT to WASH-FAST

41

Chapter 3 Evaluating the foundations that help avert antimicrobial

resistance: Performance of essential water sanitation and

hygiene functions in hospitals and requirements for action

in Kenya

77

Chapter 4 Infection prevention and control during the COVID-19

pandemic: challenges and opportunities for Kenyan public

hospitals

169

Chapter 5 Evaluating hospital performance in antibiotic stewardship

to guide action at national and local levels in a lower-

middle-income setting

203

Chapter 6 Antibiotic use in Kenyan public hospitals: Prevalence,

appropriateness and link to guideline availability

233

Chapter 7 Using treatment guidelines to improve antibiotic use:

Insights from an antibiotic point prevalence survey in

Kenyan public hospitals

291

Chapter 8 Using a common data platform to facilitate audit and

feedback on the quality of hospital care provided to sick

newborns in Kenya

305

Chapter 9 General Discussion 329

Chapter 10 Summary, Samenvatting, Acknowledgements, Author

Portfolio & Resume

349

6

ABBREVIATIONS

A&F: Audit and feedback

ABS: Antibiotic Stewardship

AMR: Antimicrobial Resistance

ATC: Anatomical Therapeutic Chemical Classification System

BPP: Basic Paediatric Protocols

CIN: Clinical Information Network

ETAT+: Emergency Triage Assessment and Treatment Plus Admission

HAI: Hospital-Acquired Infections

IQR: Interquartile range

IPC: Infection Prevention and Control

KEMRI: Kenya Medical Research Institute

LMIC: Low and middle-income countries

MOH: Ministry of Health

NAR: Neonatal Admission Record

PPS: Point Prevalence Survey

REDCap: Research Electronic Data Capture

WASH: Water Sanitation and Hygiene

WASH-FIT: Water Sanitation and Hygiene Facility Improvement Tool

WASH-FAST: Water Sanitation and Hygiene facility survey tool

WHO: World Health Organisation

7

Chapter 1: Introduction

8

BACKGROUND Antibiotic use for infections was heralded as one of the significant discoveries

of our time, saving millions of lives since the discovery of penicillin in 1928

[1]. The discovery of newer molecules has not kept pace with the growing

resistance to antibiotics that has accompanied their ever-widening use.

Antimicrobial resistance (AMR) is broadly defined as the resistance of a

pathogenic micro-organism to an antimicrobial agent which was initially

effective for the treatment of infections caused by the organism [2]. In

addition to AMR related deaths, estimated at 10 million per year by 2050 [3],

worrying social and economic effects have also been noted. These include

pushing millions of people into extreme poverty as a result of increased costs

from drug-resistant infections and reduced productivity and loss of labour

due to sickness, and therefore widening the gap between the rich and

developing nations [4]. Although the threat of AMR is global, the most

significant effects will be in the low-income countries found mainly in sub-

Saharan Africa and the Asian continent [4]. These regions are already dealing

with multiple other health challenges, and the effects of AMR are expected

to worsen these.

Interventions to combat AMR mainly involve the integration of strategies

across sectors and levels of different systems. The ‘One Health’ approach is

one good example of this integration. The organisms that affect animal and

plant health also have a direct implication on the infections in humans and

vice versa. With this realisation interventions to combat AMR have also

targeted antimicrobial use in animals and plants [5]. Within the health sector,

interventions to combat AMR have been proposed at all levels, from the

community and hospital levels to the regional and international levels. These

interventions include the development of vaccines, newer drugs, increased

9

public awareness, improved diagnostics and surveillance, rational

antimicrobial use and infection prevention and control (IPC) measures[3, 6]

(Figure 1). Other interventions include the health workers educating the

patients on antibiotic use and proper prescription of antimicrobials. There is

now a push to actively and responsibly involve patients in IPC activities since

some of these patients have the potential to transmit disease. The patients

should, therefore, be viewed as real partners in IPC [7].

Figure 1 Strategies to combat Antimicrobial resistance at community, hospital, national and

international level

Despite the acceptance of the broad impact of antibiotic resistance and the

need for a one health approach, it is essential to recognise that hospitals play

an important role in resistance. Additionally, antibiotics are so central to the

role of the modern hospital, that any diminishment of their effectiveness

affects the overall capabilities of hospitals[8]. Hence, while an integrated and

holistic multisectoral approach encompassing multiple interventions is

10

pivotal for the management of AMR, this work focusses on hospital-level

strategies. Specifically, I focus on IPC and antibiotic stewardship in inpatient

settings [9, 10].

Here, I highlight the roles played by IPC and Water Sanitation and

Hygiene(WASH) in hospitals and their contributions to AMR. Additionally, I

propose why improving antibiotic use, and stewardship are central to the

reduction of AMR.

1.1 Infection Prevention and Control and Water Sanitation and Hygiene

Infection Prevention and Control is broadly defined as “a scientific approach

and practical solution designed to prevent harm caused by infection to

patients and health workers” [11]. Some of the IPC domains include hand

hygiene, waste management and use of personal protective equipment

(PPE). Water sanitation and hygiene (WASH) tends to crosscut all the major

IPC core components. These include water, sanitation and healthcare waste

management, hand hygiene, environmental management, cleanliness and

disinfection. Proper IPC and WASH structures within hospitals are associated

with better uptake of care by the community and increased health worker

morale and efficiency [9, 12]. They are also vital in moving countries towards

universal health coverage by improved quality of care and reducing

unnecessary expenditure caused by antimicrobial resistance in hospitals [13].

The contribution of WASH to AMR has been demonstrated at the community

and hospital levels. At the hospital level, the effects of these weak WASH and

IPC structures are demonstrated by the high burden of hospital-acquired

infections (HAIs). These HAIs are the most frequent adverse events in the

provision of healthcare worldwide which are largely preventable. Multidrug-

11

resistant organisms often cause these infections, which result in extended

hospital stays, increased costs of care and are associated with increased

mortality [14]. It is estimated that 15% of hospitalised patients in developing

countries develop hospital-acquired infections [15]. These HAIs include post-

partum and neonatal infections due to unclean deliveries [16].

In addition to hospitals being the epicentre for HAI’s, hospitals generate

clinical waste which if not well handled, contaminates the environment with

drug-resistant organisms with a resultant spread of disease to communities

[17]. With a focus on the one health approach, contamination of the water

and soils with these organisms also affects animal and plant health.

Poor WASH also alters the clinicians’ behaviour and practice. Some of which

lead to the overuse of antibiotics in hospitals. Recent surveys from Uganda

and Tanzania present antibiotic prescription as a “solution” or “quick fix” to

inherent problems of poor care due to weak health systems and lack of

proper hygiene structures. This focus on prescription instead of fixing the

underlying hygiene problems and other key issues worsens AMR due to

irrational use of antibiotics [18].

Improving behaviour, the environment (including WASH) coupled with

proper use of antibiotics can markedly reduce the burden of HAI[19]. In

resource-limited settings, providing high-quality primary care, some of which

entails proper WASH and IPC could see a marked reduction in hospital

mortalities, including neonatal deaths [20].

Recent global surveys by the World Health Organisation(WHO), in 2014, show

significant shortcomings in the state of WASH across all primary care health

facilities. About 25% of the facilities assessed did not have essential water

services, and 20% of them had poor sanitation services, including inadequate

functioning toilets. More than half of the facilities had inadequate waste

12

disposal arrangements [21]. Similar but limited work has been carried out in

Kenyan hospitals accessing WASH and IPC capacity. In a recent survey

observing IPC behaviour among health workers in primary health care

facilities in Kenya, there was poor compliance to hand hygiene practices, use

of personal protective equipment including the use of gloves, poor handling

of needles, sharps and waste segregation. The study found a weak

association between compliance and the level of clinician knowledge and

availability of supplies and it also demonstrated the need for reinforcing the

behavioural norms to improve IPC in addition to the provision of supplies and

training [22].

In addition to poor behaviour, weak WASH infrastructure and poor

leadership have also been demonstrated to reduce compliance to IPC.

Additional work assessing IPC compliance among hospital workers in a

Kenyan district hospital demonstrated the significant barriers to IPC

compliance included frequent water shortages, poor leadership due to

inactive IPC committees and lack of frequent training on IPC [23].

To improve IPC, the WHO developed guidelines on the core areas of focus on

IPC at national and hospital levels [24]. These components form the platform

on which most IPC arrangements at the facility level should be built (Figure

2).

13

Figure 1 Core components of Infection prevention and control

Based on these components, the Kenyan Ministry of Health developed

guidelines on IPC. The main areas of focus include the use of personal

protective equipment, hand hygiene, prevention of needle stick injuries,

waste disposal and environmental hygiene [25]. The implementation of these

Kenyan guidelines is underway, but there has been no formal assessment of

the state of IPC in Kenyan hospitals, and no clear mechanism for audit and

feedback exists to promote the improvement of IPC in Kenya.

At a global level, based on the eight IPC core components and from the

results of the WASH global survey, the WHO developed a facility improvement

tool (WASH FIT) that expands on existing standards for safe and clean health

facilities [13]. These standards include the “Essential Environmental Health

Standards in Health Care (EEHH)” that describe standards of water quality

and quantity and health care waste disposal [26].

Using WASH assessment tools like the WASH-FIT to assess IPC infrastructure

within health facilities is a good starting point to identify the IPC gaps. Training

materials on the use of the WASH-FIT are freely available, and the tool is easy

Core components of Infection prevention and control

1. IPC programme with a dedicated and trained team in every facility

2. Develop evidence-based guidelines to reduce HAI and AMR

3. Education and training on IPC at each facility

4. Facility-based surveillance of HAI

5. Implement IPC activities using a multimodal strategy

6. Regular monitoring, audit and feedback on IPC standards

7. Monitor workload, staffing and bed occupancy

8. Clean environment, materials and equipment for patient care

14

to use within hospital settings. In this thesis, I propose the use of this WASH-

FIT tool for IPC assessment in health facilities in Kenya and other resource-

limited settings.

1.2 Antibiotic Stewardship and Rational antimicrobial use

Rational antimicrobial use is another key strategy to reduce AMR. Rational

drug use as defined by the WHO demands that “the appropriate drug be

prescribed, that it be available at the right time at a price people can afford,

that it be dispensed correctly, and that it be taken in the right dose at the

right intervals and for the right length of time. The appropriate drug must be

effective, and of acceptable quality and safety defined” [27].

Antibiotic Stewardship (ABS) is defined as “the optimal selection, dosage, and

duration of antimicrobial treatment that results in the best clinical outcome

for the treatment or prevention of infection, with minimal toxicity to the

patient and minimal impact on subsequent resistance” [28]. Efforts to

promote rational antibiotic use have seen the development of antibiotic

stewardship policies and hospital committees. These policies include clinician

education on sound prescription patterns, hospital antibiotic audits,

antibiotic use guidelines, essential drug lists and surveillance systems to track

AMR patterns [29]. These programmes have been shown to decrease

antibiotic use, HAIs and subsequently AMR [10]. In many African countries,

these stewardship activities suffer from poor leadership and lack of funding.

Additionally, guidelines may not exist, or may not be frequently updated.

Also, few surveillance programmes to help to monitor the use of antibiotics

are currently in place [30].

15

Antibiotic stewardship, prescription and use is often driven by complex,

inter-related influences spanning macro ([inter]]National), meso (Hospital)

and micro (Patient and Clinician) levels of the health system.

At the patient level, age, sex, disease severity, and comorbidity influence the

number and type of antibiotics prescribed [31, 32]. At the clinician level,

some of the factors that drive antibiotic prescriptions include fear of missing

infections, poor practices taught by senior clinicians, prescription pressure

from patients and fellow clinicians and the level of training [33-35]. To retain

the doctor-patient relationship, some clinicians report having to prescribe

antibiotics with no clear indications for example in respiratory tract infections

which may be viral in origin or ‘just in case’ a bacterial infection is present due

to diagnostic uncertainty [36, 37]. In a review of prescriptions practices by

physicians in India, some prescribed antibiotics for sustainability and

financial reasons, with the physicians reporting they prescribed because they

did not want to watch and wait and therefore lose the patients to other

clinicians due to poor outcomes [37]. Other factors related to clinicians that

affect their prescriptions include their perceptions of patient safety and their

willingness to comply with treatment guidelines [38].

At the hospital level, the level of care provided by the facility affects antibiotic

use with critical care and neonatal units having more prescriptions and

greater use of non-first line regimens [39]. In other facilities where there was

an oversupply of antibiotics with near expiry dates, the clinicians resulted in

prescribing antibiotics that may not be necessary for the patients [37]. In

hospital presence of laboratory support and clinical practice guidelines can

improve decision making and positively affect antibiotic usage [40, 41].

Antibiotic stewardship committees, infectious disease specialists and

16

educational programmes that train the clinicians on antibiotic use also

influence how antibiotics are prescribed in these facilities [33, 42].

At the national level, the provision of national treatment guidelines and

training of medical personnel and the provision of essential drug lists

influences prescriptions across all levels of care [43]. There are also

professional norms instilled on health workers that define best practice

which would influence antibiotic use and stewardship [44]. At international

levels, some of the drivers include the efforts by organisations like the WHO

to improve antibiotic stewardship.

Studies examining antimicrobial resistance, antibiotic use and stewardship in

Africa and locally point to some serious challenges. The World Health

Organization (WHO) Antimicrobial Resistance Global Report on Surveillance

(2014) reports high resistance among major disease-causing organisms

(Escherichia coli, Klebsiella pneumoniae and Staphylococcus aureus) in most

regions across the world [45]. In the African region, data on AMR have been

collated in a recent systematic review of 144 studies. More than half of these

studies were conducted after 2013 [46]. The report shows significant median

resistance (MR) estimates to commonly used antibiotics for common

illnesses that are of significant concern (e.g. 26% MR to penicillin in

Streptococcus pneumoniae; 34% MR to amoxicillin in Haemophilus

Influenzae). MR was also estimated at 88% to amoxicillin and 30% to

gentamicin in Escherichia coli, which is a common bacterial pathogen in the

African region in the neonatal populations in particular. Similarly, East African

data on antimicrobial resistance reports increased resistance of

Streptococcus pneumoniae and Haemophilus Influenzae and other gram-

negative organisms to the common first-line therapies[47]. These alarming

17

levels of resistance highlight the need for prevention of further increases

while novel interventions are developed and implemented.

Data on antibiotic use in Africa report high proportions of use of beta-lactam

penicillins and cephalosporins in adult patients with less than 15% of these

prescriptions supported by bacteriological and antibiotic susceptibility tests

[48]. Kenyan studies which are mainly from single hospitals show that more

than 70% of hospitalised patients are on antibiotic treatment. This is despite

often having conditions that do not warrant antibiotic treatment [49]. The

common antibiotic prescriptions include penicillin, 3rd generation

cephalosporins and imidazole derivatives, mainly metronidazole [50].

To improve the availability of data on antibiotics, the WHO commissioned a

global surveillance system to measure antibiotic consumption at national and

regional levels[51]. This system encourages monitoring of antibiotic use at

the community and hospital level. At the community level, this surveillance

includes primary care settings and outpatient clinics. The system monitors

hospital data at both the public and private sectors looking at the antibiotics

used (products and packages). However, it does not consider prescription

quality at, ward or patient level. To measure antibiotic use and quality of

prescriptions at the hospital and patient level, point prevalence surveys(PPS)

have been extensively used [50, 52]. The Global Point Prevalence Survey of

Antimicrobial Consumption and Resistance (Global-PPS), is one of the leading

surveys available globally. Since its inception in 2015, the Global PPS has been

conducted yearly in more than 300 hospitals globally, but very few in Africa

and Kenya, only a few private hospitals are involved. The Global-PPS aims to

generate global data on antibiotic use and resistance in inpatient settings. It

also aims to identify and measure interventions to improve quality of

prescription [48, 53]. If done at a larger scale in Kenya, these surveys would

18

provide crucial data on antibiotic use. Part of this work carries out these

surveys in select Kenyan hospitals.

Following the endorsement of the Global action plan on AMR during the 68th

World Health assembly, Kenya developed a national policy on the prevention

and containment of AMR in 2017 [54]. The main objectives of the policy are

to improve awareness on AMR, increase AMR surveillance, reduction of

infections through improved IPC and optimisation of antimicrobial use. In the

policy, the national government is tasked with the process of resource

mobilisation to meet the objectives of the policy, including AMR surveillance

and also provide resources for related activities like IPC [54]. Little is known of

how these and other related policies have affected antibiotic use in Kenyan

public hospitals, and this work seeks to understand this.

1.3 IPC and AMR in the face of emerging infectious diseases

IPC and WASH are essential in the management of pandemics, including

COVID 19 and Ebola. Proper hand hygiene is crucial in reducing virus spread

[55]. In the hospital setting, waste disposal, disinfection, and the proper use

of personal protective equipment (PPE) are key to mitigate the spread of

COVID-19 and protect health workers [56].

The emergence of COVID-19 has also been a significant threat to AMR. This is

due to the increased use of antibiotics due to the fear of possible secondary

bacterial infections in COVID-19 patients [57]. Hospitalisations, including

those due to COVID-19 infection increase the risk of HAI, some of which are

drug-resistant infections [14]. The increased use of bactericidal cleaning

agents and disinfectants may increase the selection of drug-resistant genes

in hospitals and communities [58]. The disruption of health services due to

COVID-19 may result in a surge in vaccine-preventable diseases and HIV

19

associated infections which may mean increased antibiotic use further

driving AMR [59].

1.4 Kenyan Health System

In Kenya, hospitals are categorised based on the care provided into six levels.

The lowest level (level 1) is the community health services; level 2,

dispensaries and clinics; level 3 is health centres; level 4 are sub-county

hospitals (formerly district hospitals) and some private hospitals; level 5 is the

county referral hospitals and larger private hospitals, and the highest level

(level 6) is the national referral hospitals [60]. In August 2010, Kenya adopted

a devolved system of government with the creation of 47 counties and

devolved the management of health services to the county governments [61].

The central government is responsible for policy generation and the

management of level 6 hospitals. It also retained the running of some critical

programmes including national tuberculosis, HIV, blood transfusion and

immunisation programmes. Provision of curative services, the running of

hospitals (level 1 - 5) and human resource management are under the county

governments with funding by the central government. Funding for health in

Kenya is mainly through the government and donor agencies, contributions to

the national hospital insurance funds (NHIF) and private medical insurances

and out of pocket payments [62].

Overall, the public facilities account for 52% of all facilities, private facilities

at 37% and faith-based facilities at 11% [63]. The public health facilities cater

mainly for those in the lower-income brackets, most of whom do not have

medical insurance and pay for care out of pocket. Out of pocket expenditure

accounts for about 30% of the total health expenditure in Kenya [64]. Only

20

about 20% of Kenyans have some form of health insurance [64]. The health

insurance coverage is also skewed towards those who are wealthy. Insurance

coverage is higher in those in the highest wealth quintile at 42% coverage

compared to 3% coverage among those in the lowest quintile [65].

1.5 Study sites

This work was conducted in 16 hospitals in Kenya. These hospitals are located

in 11 counties in Kenya. Fifteen of these hospitals are selected purposively as

they are part of the clinical information network (CIN). The CIN is a

collaboration between the Ministry of Health, county departments of health,

Kenya Paediatric Association and the KEMRI Wellcome Trust Research

Programme. The main aim of the CIN, which was established in October 2013

is to improve the quality of paediatric and neonatal care through improving

the clinical documentation and data utilisation for decision making [66]. The

hospitals in the CIN are located in the high and low malaria zones of Kenya;

they are mainly level four and five public hospitals. These hospitals provide

inpatient medical, surgical and neonatal care. Some have specialised eye,

psychiatry, neurosurgery, renal and critical care units. The bed capacity

ranges between 130- 600 beds and the number of consultants in these

facilities ranging between 5-26. The additional hospital (H15 in figure 3

below) is a national referral hospital with 50 wards and 1800 beds. For this

hospital, due to logistical reasons, we conducted our study only in the

neonatal units. Figure 3 below provides the geographical location of these

hospitals in Kenya

21

Figure 3: Geographical location of the 16 study hospitals. The five hospitals in western

Kenya are located in the high malaria regions of Kenya.

2 JUSTIFICATION

As discussed above, I believe IPC and ABS are interlinked and understanding

them both individually and together is essential to combating AMR.

Unfortunately, little is known about their current level of implementation

across different settings in Kenya, what explains variability, and what

policymakers should prioritise. Additionally, this work seeks to identify what

22

gaps need to be addressed as starting points to strengthening these two

linked strategies to combat AMR.

To generate the contextual understanding of IPC arrangements in Kenya, this

work involved re-designing existing WHO assessment tools (WASH-FIT) to

make it appropriate for use in the context of larger hospitals that have

inpatient units [13]. This modified WASH-FIT tool was then used to assess

infrastructural, material and human resources to support WASH services,

across 15 county hospitals (mainly level 4 and 5 hospitals) in Kenya. Realising

the crucial role played by IPC and WASH in the containment of COVID-19, I

conducted follow-up interviews in the survey hospitals after the emergence

of COVID-19 in Kenya.

While most antibiotics are prescribed in outpatient clinics, consideration of

inpatient settings is crucial. Inpatient wards are where severe infections are

typically treated, and they also tend to be the sites where most HAI’s are

acquired and thus, are where patients are most likely to suffer the worst

clinical outcomes from AMR. Hospital inpatient areas also typically serve as

training centres where prescribing habits and behaviours often develop. This

work, therefore, focused on antibiotic use in inpatient settings. The work

examined antibiotic use and guideline availability across service units in

Kenyan public hospitals.

Additionally, I assessed the organisation of antibiotic stewardship structures

across these hospitals to highlight the challenges and opportunities to

improve antibiotic stewardship and use. Lastly, I propose some context-

relevant interventions specific to antibiotic stewardship and use. In

particular, I highlight the need for clinical practice guidelines in hospitals to

improve antibiotic usage. I also explore how improvements in data systems

might support better, continuous evaluation in the future.

23

OBJECTIVES

Main Objective

To assess hospitals Infection Prevention and Control (IPC) and Antibiotic

Stewardship(ABS) capacity as part of tracking and tackling efforts to limit

antimicrobial resistance in Kenya.

Specific Objectives

1. Develop a survey tool that can be applied at the national or sub-national

level to monitor WASH and infection prevention and control

performance in hospitals.

2. To evaluate the WASH and Infection prevention and control

arrangements in Kenyan county hospitals and explore how these may

vary within a single public health system.

3. To evaluate the opportunities and challenges for IPC and WASH

implementation in the COVID-19 pandemic.

4. To evaluate antibiotic stewardship arrangements in Kenyan hospitals to

guide action at the regional and national level.

5. To examine patterns of antibiotic use and guideline availability across

hospitals and medical specialities in Kenyan county hospitals.

6. To make a case for further development and use of treatment guidelines

in improving antibiotic usage in hospitals in low and middle-income

countries.

7. To examine the feasibility of creating an inpatient data platform to

support the regular assessment of appropriate and correct treatment in

Kenyan hospitals, taking inpatient neonatal units as an example.

24

3 STUDY DESIGN

Objective one took a mixed-methods approach. The development of the

survey tool involved conceptual thinking about the nature of the tool and its

purpose in the re-design phase, it also included a process of document review

and interviews with managers and a second phase was a consensus meeting

with relevant stakeholders to validate/critique the proposed tool

modifications. A mix of cross-sectional surveys and in-depth interviews were

employed for objective two and four. In-depth phone interviews were also

used for objective three to evaluate the opportunities and challenges for IPC

during the COVID-19 pandemic. A cross-sectional survey and hierarchical

modelling approaches were used to examine the factors influencing

antibiotic use and guideline availability (objective five). Objective six used

some data from the antibiotic survey to describe the importance of

developing clinical treatment guidelines. For the description of the common

data platform (objective seven), the data analysed were part of routine

admission data collected over three years in a neonatal unit, under the

neonatal clinical information network.

4 OUTLINE OF THE THESIS

This thesis consists of three main parts;

Part 1 ( Chapter two, three and four): This section assesses the structural and

organisational capacity of Infection prevention and control in Kenyan public

hospitals using a modification of the WHO WASH- FIT. To do this, I adapted the

WASH-FIT to a Water Sanitation and Hygiene Facility Survey Tool (WASH-

FAST) to enable assessment in larger hospitals and also assign responsibility

for action at different levels of the health systems.

25

Using the WASH-FAST, an assessment was carried out in 14 hospitals to

establish the status of WASH within hospital wards and across different

hospitals. These surveys were complemented by in-depth interviews with key

stakeholders at the hospital (hospital managers, doctors and nursed) and

county level to generate a contextual understanding of the survey results.

As follow up to the original surveys, I conducted follow up phone interviews

to the health workers in 11 hospitals during the COVID-19 pandemic to

understand if the findings and actions from the initial surveys had improved

preparedness and establish the new challenges the COVID-19 pandemic had

brought to the hospitals. Lastly, using a commentary, I present some insights

on the opportunities the COVID-19 pandemic brings to improve IPC.

This section concludes with two additions; an opinion piece on the role of the

infection prevention and control committees in the leadership of IPC within

the hospitals and why we should focus on strengthening these committees.

The second addendum is a policy brief that was prepared and presented to

the Ministry of Health and other stakeholders in Kenya. The brief provides an

overview of the WASH/IPC work.

Part 2 (Chapter five and six): This section focusses on antibiotic stewardship

and use. Using a set of 17 indicators, we assessed antibiotic stewardship

capacities in the study hospitals. These surveys were accompanied with in-

depth interviews with health workers, including pharmacy and laboratory

staff, to understand the stewardship practices better. Using a point

prevalence survey, I studied antibiotic use among hospitalised patients on

antibiotic treatment in the survey hospitals. The point prevalence survey also

assessed guideline availability within the hospital departments. Using

predefined criteria, I also established the appropriateness of antibiotic

treatment for these patients based on the documented diagnosis.

26

Part 3 (Chapter seven and eight): Based on findings from the first two parts,

I propose some interventions that would improve IPC and antibiotic

stewardship. The first is the importance of developing clinical practice

guidelines for use in hospitals. I use data from the point prevalence survey to

show the effects of having guidelines on treatment appropriateness. I

suggest some common clinical conditions in urgent need of guidelines and

highlight the guideline development process in resource-limited settings.

Secondly, I use routine clinical data from a large neonatal unit which was part

of the survey to illustrate how good data can improve decision making. Here

I illustrate how creating a standardised neonatal record together with cycles

of audit and feedback (A&F) to improve clinical documentation can generate

high-quality data and improve antibiotic dosing accuracy. Figure 4 below

illustrates how these sections are linked;

27

Figure 4; Framework of work to understand and improve the IPC and ABS arrangements

in Kenyan Hospitals

5 LIST OF PUBLICATIONS AND AUTHOR CONTRIBUTIONS

The individual papers and the authors are indicated in every chapter of this

thesis. The full table with author contributions is in the supplementary

materials (S1) which provides a list of publications by the objectives and

indicates the roles and contributions of all the authors.

6 FUNDING INFORMATION

This work was supported by funds from the economic and social research

council ESRCS # ES/P004938/1, and a Senior Research Fellowship awarded to

Prof Mike English by The Wellcome Trust (#207522). Michuki Maina is

supported by a grant from by the Initiative to Develop African Research

Leaders (IDeAL) through the DELTAS Africa Initiative [DEL-15-003], an

independent funding scheme of the African Academy of Sciences (AAS) ’s

28

Alliance for Accelerating Excellence in Science in Africa (AESA) and supported

by the New Partnership for Africa’s Development Planning and Coordinating

Agency (NEPAD Agency) with funding from the Wellcome Trust

[107769/Z/10/Z] and the UK government. The funders had no role in the

preparation or submission of this work.

29

7 REFERENCES

1. Ventola, C.L., The Antibiotic Resistance Crisis: Part 1: Causes and Threats. P t, 2015.

40(4): p. 277-83.

2. World Health Organization. Antimicrobial resistance. 2017 [cited 2017 18th

December]; Available from:http://www.who.int/mediacentre/factsheets/fs194/en/.

3. O’Neil, J., Review on Antimicrobial Resistance: Tackling Drug-Resistant Infections

Globally: Final Report and Recommendations. Wellcome Trust: London, UK, 2016.

4. Jonas, O.B., et al., Drug-resistant infections: a threat to our economic future. World

Bank Group, Washington, DC, 2017.

5. White, A. and J.M. Hughes, Critical Importance of a One Health Approach to

Antimicrobial Resistance. EcoHealth, 2019. 16(3): p. 404-409.

6. Rajesh R Uchil, G.S.K., Vijay M Katekhaye, Onkar C Swami, Strategies to Combat

Antimicrobial Resistance. Journal of Clinical and Diagnostic Research, 2014. 8(7): p.

ME01-ME04.

7. Fernandes Agreli, H., et al., Patient involvement in the implementation of infection

prevention and control guidelines and associated interventions: a scoping review. BMJ

open, 2019. 9(3): p. e025824- e025824.

8. Chandler, C.I., Current accounts of antimicrobial resistance: stabilisation,

individualisation and antibiotics as infrastructure. Palgrave communications, 2019.

5(1): p. 1-13.

9. Rainey R and Weinger M. The Role of Water, Sanitation and Hygiene (WASH) in

Healthcare Settings to Reduce Transmission of Antimicrobial Resistance. . AMR Control

2016 [cited 2017 18th December]; Available from:

http://resistancecontrol.info/infection-prevention-and-control/the-role-of-water-

sanitation-and-hygiene-wash-in-healthcare-settings-to-reduce-transmission-of-

antimicrobial- resistance/.

10. Paterson, D.L., The Role of Antimicrobial Management Programs in Optimizing

Antibiotic Prescribing within Hospitals. Clinical Infectious Diseases, 2006.

42(Supplement_2): p. S90-S95.

11. World Health Organisation. Infection prevention and control. 2018 [cited 2018 25th

October]; Available from: http://www.who.int/infection-prevention/about/ipc/en/.

12. World Health Organization, WASH in health care facilities: Links with the network for

improving quality of care for maternal newborn and child health 2017.

13. World Health Organization and UNICEF, Water and Sanitation for Health Facility

Improvement Tool (WASH FIT): a practical guide for improving quality of care through

water, sanitation and hygiene in health care facilities. 2017: Geneva.

14. Ducel, G., et al., Prevention of hospital-acquired infections: a practical guide. 2002.

15. Allegranzi, B., et al., Burden of endemic health-care-associated infection in developing

countries: systematic review and meta-analysis. The Lancet, 2011. 377(9761): p. 228-

241.

16. Say, L., et al., Global causes of maternal death: a WHO systematic analysis. Lancet Glob

Health, 2014.2(6): p. e323-33.

30

17. World Health Organisation. Safe management of waste from health care activities

2014 [cited 2018 18th December ]; Available from:

http://www.who.int/water_sanitation_health/publications/wastemanag/en/

18. Denyer Willis, L. and C. Chandler, Quick fix for care, productivity, hygiene and

inequality: reframing the entrenched problem of antibiotic overuse. BMJ Global Health,

2019. 4(4): p. e001590.

19. McLaws, M.-L., The relationship between hand hygiene and health care-associated

infection: it’s complicated. Infection and drug resistance, 2015. 8: p. 7.

20. Dickson, K.E., et al., Every Newborn: health-systems bottlenecks and strategies to

accelerate scale-up in countries. The Lancet, 2014. 384(9941): p. 438-454.

21. World Health Organization and UNICEF, Water, sanitation and hygiene in health care

facilities: status in low and middle income countries and way forward. 2015.

22. Bedoya, G., et al., Observations of infection prevention and control practices in primary

health care, Kenya. Bulletin of the World Health Organization, 2017. 95(7): p. 503.

23. Gichuhi, A., et al., Health Care Workers Adherence to Infection Prevention Practices

and Control Measures: A Case of a Level Four District Hospital in Kenya. American

Journal of Nursing Science, 2015. 4: p. 39-44.

24. World Health Organization, Guidelines on core components of infection prevention and

control programmes at the national and acute health care facility level. 2016: World

Health Organization.

25. Ministry of Public Health and Sanitation and Ministry of Medical Services, National

Infection Prevention and Control Guidelines for Health Care Services in Kenya. 2010,

Republic of Kenya: Nairobi.

26. John, A., B. Jamie, and C. Yves, Essential environmental health standards in health care.

Geneva: World Health Organization, 2008.

27. World Health Organization, The Rational Use of Drugs - Report of the Conference of

Experts, Nairobi 25-29 November 1985,. 1987, World Health Organization: Geneva.

28. Gerding, D.N., The search for good antimicrobial stewardship. Jt Comm J Qual Improv,

2001. 27(8): p. 403-4.

29. Centers for Disease Control and Prevention, Core elements of hospital antibiotic

stewardship programs. 2014.

30. Cox, J.A., et al., Antibiotic stewardship in low- and middle-income countries: the same

but different? Clin Microbiol Infect, 2017. 23(11): p. 812-818.

31. Monteiro, L.G.S., et al., Determinants of antibiotic prescription in paediatric patients:

The case of two hospitals in Maputo, Mozambique. South African Journal of Child

Health, 2017. 11: p. 109+.

32. Nobili, A., et al., Polypharmacy, length of hospital stay, and in-hospital mortality among

elderly patients in internal medicine wards. The REPOSI study. European Journal of

Clinical Pharmacology, 2011. 67(5): p. 507-519.

33. De Souza, V., et al., A qualitative study of factors influencing antimicrobial prescribing

by non- consultant hospital doctors. J Antimicrob Chemother, 2006. 58(4): p. 840-3.

34. Saleem, Z., et al., Antimicrobial prescribing and determinants of antimicrobial

resistance: a qualitative study among physicians in Pakistan. Int J Clin Pharm, 2019.

41(5): p. 1348-1358.

31

35. Steinke, D.T., et al., Practice factors that influence antibiotic prescribing in general

practice in Tayside. Journal of Antimicrobial Chemotherapy, 2000. 46(3): p. 509-512.

36. Butler, C.C., et al., Understanding the culture of prescribing: qualitative study of general

practitioners’ and patients’ perceptions of antibiotics for sore throats. Bmj, 1998.

317(7159): p. 637-42.

37. Kotwani, A., et al., Factors influencing primary care physicians to prescribe antibiotics

in Delhi, India. Family Practice, 2010. 27(6): p. 684-690.

38. Lucet, J.-C., et al., Antibiotic use: knowledge and perceptions in two university

hospitals. Journal of Antimicrobial Chemotherapy, 2011. 66(4): p. 936-940.

39. Benko, R., et al., Variations and determinants of antibiotic consumption in Hungarian

adult intensive care units. Pharmacoepidemiol Drug Saf, 2012. 21(1): p. 104-9.

40. Haldrup, S., et al., Microbiological point of care testing before antibiotic prescribing in

primary care: considerable variations between practices. BMC Fam Pract, 2017. 18(1):

p. 9.

41. Tan, T.Y., et al., Laboratory antibiotic susceptibility reporting and antibiotic prescribing

in general practice. Journal of Antimicrobial Chemotherapy, 2003. 51(2): p. 379-384.

42. Malani, A.N., et al., Clinical and economic outcomes from a community hospital’s

antimicrobial stewardship program. Am J Infect Control, 2013. 41(2): p. 145-8.

43. English, M., et al., Adoption of recommended practices and basic technologies in a low-

income setting. Arch Dis Child, 2014. 99(5): p. 452-6.

44. Ministry of Health, Kenya Quality Assurance Model for Health (KQAMH),Quality

Standards for Kenya Essential Package of Health. . 2009: Nairobi.

45. World Health Organisation, Antimicrobial Resistance Global Report on surveillance

2014, World Health Organization: Geneva

46. Tadesse, B.T., et al., Antimicrobial resistance in Africa: a systematic review. BMC

Infectious Diseases, 2017. 17(1): p. 616.

47. Ampaire, L., et al., A review of antimicrobial resistance in East Africa. Afr J Lab Med,

2016. 5(1): p. 432.

48. Versporten, A., et al., Antimicrobial consumption and resistance in adult hospital

inpatients in 53 countries: results of an internet-based global point prevalence survey.

The Lancet Global Health, 2018. 6(6): p. e619-e629.

49. Mulwa, N.C., et al., Patterns of Prescribing Practices in Makueni County Referral

Hospital, Kenya. African Journal of Pharmacology and Therapeutics, 2015. 4(4).

50. Okoth, C., et al., Point prevalence survey of antibiotic use and resistance at a referral

hospital in Kenya: findings and implications. Hospital practice, 2018. 46(3): p. 128-136.

51. World Health Organisation, WHO methodology for a global programme on surveillance

of antimicrobial consumption. 2016, WHO, Geneva.

52. Gharbi, M., et al., Using a simple point-prevalence survey to define appropriate

antibiotic prescribing in hospitalised children across the UK. BMJ Open, 2016. 6(11).

53. GLOBAL-PPS. Global Point Prevalence Survey of Antimicrobial Consumption and

Resistance 2015 [cited 2017 16th February]; Available from: http://www.global-

pps.com/.

54. Republic of Kenya, National Policy for the Prevention and Containment of Antimicrobial

Resistance. 2017: Nairobi, Kenya.

32

55. World Health Organization, Water, sanitation, hygiene, and waste management for the

COVID-19 virus: interim guidance, 23 April 2020. 2020, World Health Organization.

56. Herron, J.B.T., et al., Personal protective equipment and COVID 19- a risk to healthcare

staff? Br J Oral Maxillofac Surg, 2020.

57. Zhou, F., et al., Clinical course and risk factors for mortality of adult inpatients with

COVID-19 in Wuhan, China: a retrospective cohort study. Lancet, 2020. 395(10229): p.

1054-1062.

58. Caselli, E., Hygiene: microbial strategies to reduce pathogens and drug resistance in

clinical settings. Microbial Biotechnology, 2017. 10(5): p. 1079-1083.

59. Getahun, H., et al., Tackling antimicrobial resistance in the COVID-19 pandemic. Bull

World Health Organ, 2020. 98: p. 442-442A.

60. World Health Organization, Primary Health Care Systems (PRIMASYS): Case study from

Kenya. 2017, Geneva.

61. Government of Kenya, The constitution of Kenya. Government Printer. Kenya: Nairobi,

2010.

62. Barasa, E., P. Nguhiu, and D. McIntyre, Measuring progress towards sustainable

development goal 3.8 on universal health coverage in Kenya. BMJ global health, 2018.

3(3).

63. PharmAccess Group. A closer look at the healthcare system in Kenya. 2016 [cited 2020

26th August ]; Available from: https://www.pharmaccess.org/wp-

content/uploads/2018/01/The-healthcare-system- in-Kenya.pdf.

64. Dutta, A., et al., Kenya Health Financing System Assessment, 2018: Time to Pick the

Best Path. 2018, Palladium: Washington DC.

65. Ministry of Health, Kenya Household Health Expenditure and Utilisation Survey. 2018:

Nairobi.

66. Tuti, T., et al., Improving documentation of clinical care within a clinical information

network: an essential initial step in efforts to understand and improve care in Kenyan

hospitals. BMJ Global Health, 2016. 1(1).

33

Supplementary Materials S1: List of Manuscripts and Author Contributions

Chapter Manuscript Title Authors Author Roles

2 Extending the use of the World Health

Organisations’ water sanitation and hygiene

assessment tool for surveys in hospitals - from

WASH-FIT to WASH- FAST. PLoS One. 2019;14(12):

e0226548.

https://doi.org/10.1371/journal.pone.0226548

Michuki Maina, Olga

Tosas-Auguet, Jacob

McKnight, Mathias

Zosi, Grace Kimemia,

Paul Mwaniki,

Arabella Hayter,

Margaret

Montgomery,

Constance Schultsz

and Mike English

Michuki Maina: Conceptualisation, Data curation, Formal

analysis, Investigation, Methodology, Visualisation,

Writing – original draft, Writing

Olga Tosas-Auguet: Conceptualisation, Formal analysis,

Funding acquisition, Investigation, Methodology,

Supervision, Writing

Jacob McKnight: Conceptualisation, Formal analysis,

Investigation, Methodology, Supervision, Validation,

Writing

Mathias Zosi: Data curation, Investigation, Project

administration, Writing – review & editing

Grace Kimemia/ Paul Mwaniki: Formal analysis,

Investigation, Methodology, Software, Writing – review

& editing.

Arabella Hayter/ Margaret Montgomery: Methodology,

Resources, Validation, Writing – review & editing.

Constance Schultsz: Conceptualisation, Methodology,

Supervision, Writing

Mike English: Conceptualisation, Funding acquisition,

Methodology,Project administration, Supervision,

Validation, Writing – review & editing

34

3 Evaluating the foundations that help avert

antimicrobial resistance: Performance of essential

water sanitation and hygiene functions in hospitals

and requirements for action in Kenya. PLoS One.

2019;14(10): e0222922.

https://doi.org/10.1371/journal.pone.0222922

Michuki Maina, Olga

Tosas-Auguet, Jacob

McKnight, Mathias

Zosi, Grace Kimemia,

Paul Mwaniki,

Constance Schultsz,

Mike English

Michuki Maina: Conceptualisation, Data curation, Formal

analysis, Investigation, Methodology, Project

administration, Software, Supervision, Validation,

Visualization, Writing – original draft, Writing review &

editing

Olga Tosas-Auguet: Conceptualisation, Formal analysis,

Funding acquisition, Investigation, Methodology, Project

administration, Supervision, Validation, Visualization,

Writing – review & editing Jacob McKnight: Formal

analysis, Investigation, Methodology, Supervision,

Validation, review.

Mathias Zosi: Data curation, Investigation, Project

administration, Supervision, Writing – review & editing

Grace Kimemia/ Paul Mwaniki Formal analysis,

Investigation, Software, Validation, Writing – review &

editing

Constance Schultsz: Conceptualisation, Methodology,

Supervision, Writing – review & editing

Mike English: Conceptualisation, Funding acquisition,

Project

administration, Supervision, Writing – review & editing

35

4 Maina M, Tosas-Auguet O,

English M et al. Infection

prevention and control

during the COVID-19

pandemic: challenges and

opportunities for Kenyan

public hospitals [version 1;

peer review: 2

approved]. Wellcome Open

Res 2020, 5:211

(https://doi.org/10.12688/we

llcomeopenres.16222.1)

Michuki Maina, Olga

Tosas- Auguet, Mike

English, Constance

Schultsz and Jacob

McKnight

Michuki Maina: Conceptualisation, Interviews, Formal analysis,

interpretation Writing – original draft, Writing – review &

editing Olga Tosas-Auguet: Conceptualisation, Supervision,

Writing – review & editing

Mike English: Conceptualisation, Funding acquisition,

Project administration, interpretation, Writing – review

& editing, Supervision, Writing – review & editing

Constance Schultsz: Conceptualisation, Project

administration, interpretation, Writing – review & editing,

Supervision, Writing – review & editing

Jacob McKnight: Conceptualisation, Formal analysis,

Methodology, interpretation, Supervision, Writing – original

draft, Writing – review & editing

4(Addendum) COVID-19: An opportunity to

improve infection prevention and

control in LMICs

The Lancet Global Health,

Access Published:

August 13, 2020, DOI:

https://doi.org/10.1016

/S2214-109X(20)30352-

Michuki Maina, Olga

Tosas-Auguet, Mike

English, Constance

Schultsz, Jacob McKnight

Michuki Maina: Conceptualisation, Interviews, Formal

analysis, interpretation, Writing – original draft, review &

editing

Mike English/ Constance Schultsz/ Olga Tosas/ Jacob McKnight:

Conceptualisation, Methodology, interpretation Supervision,

Writing – review & editing

36

4(Addendum) Performance in water sanitation

and hygiene and infection

prevention and control in

Kenyan hospitals and relevance

for the COVID-19 pandemic.

2020, KEMRI Wellcome Trust

Research Programme.

(Policy Brief)

Michuki Maina, Olga

Tosas-Auguet, Jacob

McKnight, Mathias Zosi,

Grace Kimemia, Paul

Mwaniki, Constance

Schultsz, Mike English

Michuki Maina: Conceptualisation, Data curation, Formal

analysis, Investigation, Methodology, Visualisation, Writing –

original draft, Writing – review & editing

Olga Tosas-Auguet: Conceptualisation, Formal analysis,

Funding acquisition, Investigation, Methodology, Project

administration, Supervision, Validation, Visualization, Writing

– review & editing Jacob McKnight: Formal analysis,

Investigation, Methodology, Supervision, Validation,

Visualization, Writing – review & editing Mathias Zosi: Data

curation, Investigation, Project administration, Supervision,

Writing – review & editing

Grace Kimemia/ Paul Mwaniki: Formal analysis, Investigation,

Software, Validation, Writing – review & editing

Constance Schultsz: Conceptualisation, Methodology, Supervision,

Writing – review & editing

Mike English: Conceptualisation, Funding acquisition, Project

administration, Supervision, Writing – review & editing

37

5 Evaluating hospital performance

in antibiotic stewardship to

guide action at national and local

levels in a lower-middle-income

setting. Global health action.

2019;12(sup1):1761657.

https://doi.org/10.1080/16549716.

2020.1761657

Jacob McKnight, Michuki

Maina, Mathias Zosi,

Grace Kimemia, Truphena

Onyango, Constance

Schultsz, Mike English

and Olga Tosas-Auguet.

Jacob McKnight: Conceptualisation, design, acquisition of data

(interviewing), analysis and interpretation, drafting, revisions

Michuki Maina: Conceptualisation, design, acquisition of data

(interviewing, survey), analysis and interpretation, drafting,

revisions Mathias Zosi: Acquisition of data (survey), analysis and

interpretation Grace Kimemia: Acquisition of data

(interviewing), analysis and interpretation, drafting

Truphena Onyango: data acquisition(interviewing)

Paul Mwaniki: Analysis and interpretation, drafting;

Constance Schultsz/ Mike English: Conceptualisation, design,

analysis and interpretation, drafting, revisions

Olga Tosas-Auguet: Conceptualisation, design, acquisition of data

(survey), analysis and interpretation; drafting, revisions

6 Antibiotic use in Kenyan

public hospitals: Prevalence,

appropriateness and link to

guideline availability.

International Journal of infectious

diseases 2020. 99: p. 10-

18.

https://doi.org/10.1016/j.ijid.2020.

07.084

Michuki Maina, Paul

Mwaniki, Edwin

Odira, Nduku Kiko,

Jacob McKnight,

Constance Schultsz

Mike English and Olga

Tosas- Auguet

Michuki Maina: Conceptualisation, Project administration, data

collection and curation, Formal analysis, Investigation,

Methodology, Interpretation, Visualisation, Writing – original

draft, Writing – review & editing

Paul Mwaniki: Formal analysis, interpretation, visualisation,

Writing original draft, Writing – review & editing

Edwin Odira/ Nduku Kiko: Interpretation, Writing – original

draft, Supervision, Writing – review & editing

Jacob McKnight/ Constance Schultsz/ Mike English/ Olga Tosas:

Conceptualisation: Interpretation, Writing – original draft,

Supervision, Writing – review & editing

38

7 Using treatment guidelines

to improve antibiotic use:

Insights from an antibiotic

point prevalence survey in

Kenyan public hospitals

(Accepted for publication in BMJ

Global Health)

Michuki Maina, Jacob

McKnight, Olga Tosas-

Auguet, Constance

Schultsz and Mike

English

Michuki Maina: Conceptualisation, data collection, Formal

analysis, Investigation, Methodology, Interpretation,

Visualisation, Writing – original draft, Writing – review &

editing

Jacob McKnight/ Olga Tosas/ Constance Schultsz/Mike

English : Conceptualisation: Interpretation, Supervision,

Writing – review & editing

8 Using a common data platform

to facilitate audit and feedback

on the quality of hospital care

provided to sick newborns in

Kenya. BMJ Glob Health.

2018;3(5): e001027

http://dx.doi.org/10.1136/bmjgh

-2018-001027

Michuki Maina,

Jalemba Aluvaala,

Paul Mwaniki,

Olga Tosas-

Auguet,

Catherine

Mutinda, Beth

Maina, Constance Schultsz,

Mike English

Michuki Maina: Conceptualisation: Formal analysis,

Investigation, Interpretation, Visualisation, Writing – original

draft, Writing – review & editing

Jalemba Aluvaala: Conceptualisation, Interpretation, Writing –

review & editing. Paul Mwaniki: Formal analysis,

Visualisation, Writing – review & editing

Catherine Mutinda/ Beth Maina Interpretation, Writing –

review & editing

Constance Schultsz/ Olga Tosas/Mike English Interpretation,

Supervision Writing – review & editing

39

Section 1

Infection Prevention and Control in Kenyan Hospitals

40

41

Chapter 2

Extending the use of the World Health Organisations’ water sanitation and hygiene assessment tool for surveys in

hospitals - from WASH-FIT to WASH-FAST

Michuki Maina, Olga Tosas-Auguet, Jacob McKnight, Mathias Zosi,

Grace Kimemia, Paul Mwaniki, Arabella Hayter, Margaret Montgomery,

Constance Schultsz and Mike English

PLoS One. 2019;14(12): e0226548

https://doi.org/10.1371/journal.pone.0226548

42

Abstract

Background

Poor water sanitation and hygiene (WASH) in health care facilities increases

hospital-associated infections, and the resulting greater use of second-line

antibiotics drives antimicrobial resistance. Recognising the existing gaps, the

World Health Organisations’ Water and Sanitation for Health Facility

Improvement Tool (WASH-FIT) was designed for self-assessment. The tool was

designed for small primary care facilities, mainly providing outpatient and

limited inpatient care and was not designed to compare hospital performance.

Together with technical experts, we worked to adapt the tool for use in larger

facilities with multiple inpatient units (wards), allowing for comparison

between facilities and prompt action at different levels of the health system.

Methods

We adapted the existing facility improvement tool (WASH-FIT) to create a

simple numeric scoring approach. This is to illustrate the variation across

hospitals and to facilitate monitoring of progress over time and to group

indicators that can be used to identify this variation. Working with

stakeholders, we identified those responsible for action to improve WASH at

different levels of the health system and used piloting, analysis of interview

data to establish the feasibility and potential value of the WASH Facility Survey

Tool (WASH-FAST) to demonstrate such variability.

Results

We present an aggregate percentage score based on 65 indicators at the facility

level to summarise hospitals’ overall WASH status and how this varies. Thirty-

43

four of the 65 indicators spanning four WASH domains can be assessed at ward

level enabling within hospital variations to be highlighted. Three levels of

responsibility for WASH service monitoring and improvement were identified

with stakeholders: the county/regional level, senior hospital management and

hospital infection prevention and control committees.

Conclusion

We propose WASH-FAST can be used as a survey tool to assess, measure and

monitor the progress of WASH in hospitals in resource-limited settings,

providing useful data for decision making and tracking improvements over

time.

44

Introduction

Improving water supply, hygiene, sanitation and health care waste

management (segregation, collection, disposal and treatment of health care

waste) collectively abbreviated as WASH is a significant focus of the sustainable

development goals and the global health agenda [1]. In health care facilities,

this improvement is linked to specific benefits. These include reductions in

hospital-associated infections, antimicrobial resistance, better management

and control of disease outbreaks, improved staff morale and an overall

reduction in health care costs [2] [3]. The improvements in WASH also have a

positive influence at the community level as health staff model proper hygiene

practices even at the community level [4] - and may improve patients’ trust in

and experience of care and subsequently their satisfaction with and uptake of

health services [5] [6].

Gains to improve WASH in health care facilities, mainly in the low- and middle-

income countries have been slow in the last decade. In 2015, the World Health

Organization (WHO) and the United Nations Children’s Fund (UNICEF) through

the Joint Monitoring Programme for Water Supply, Sanitation and Hygiene

highlighted some the current gaps with WASH. From this evaluation of about

60,00 health facilities, almost 40% of these health care facilities did not have

access to an improved water source, about a third of them also lacked water

and soap for handwashing, and more than half lacked arrangements for safe

disposal of health care waste [7]. A majority of these facilities assessed were in

low- and middle-income countries.

In response to these challenges, the WHO/UNICEF developed the WASH in

health care facilities global action plan to “achieve universal access to WASH in

all facilities in all settings by 2030” [4]. As part of this initiative, core and

45

extended indicators to track and improve WASH in health care facilities were

developed, revised and tested across several African and Asian countries. The

”Water and Sanitation for Health Facility Improvement Tool” (WASH-FIT), which

contains these indicators, was then developed [8]. This tool has been validated

in several countries and was developed through consultation with experts and

stakeholders before its eventual roll out[9]. It is mainly targeted at facilities in

resource-limited settings. WASH-FIT covers four broad domains (Fig 1) and

comprises 65 indicators and targets for achieving minimum standards for

maintaining a safe and clean environment. These minimum standards are as set

out in the WHO Essential environmental health standards in health care[10] and

the WHO guidelines on core components of infection prevention and control

programmes at the national and acute health care facility level [11].

Fig 1 Domains assessed in WASH FIT[8]

WASH FIT was not designed for national or regional level situation analysis,

monitoring or tracking of WASH in health care facilities. Instead, the tool guides

health care facilities staff through a continuous cycle of assessing and

prioritising risks linked to poor WASH, defining and implementing

improvements and continually monitoring progress locally and autonomously.

46

WASH FIT thus focuses on actions involving maintenance and repair as well as

infrastructural and behavioural change, which are ideally integrated into

broader quality improvement plans.

WASH FIT is meant to be adaptable to the local context but was initially

developed for use in relatively small/less complex primary health care facilities

providing outpatient services, family planning, antenatal care and maternal,

newborn and child health services (including uncomplicated delivery; e.g.

health centres, health posts and small district hospitals). Following the

inspection of the facility as a whole, WASH FIT involves scoring all 65 indicators

using a three-level qualitative system (meets, partially meets, or does not meet

the required standard), but it does not generate an overall hospital score nor

can be used to generate a score for a particular service area or WASH domain.

Larger facilities (e.g. referral hospitals), however, raise specific issues. They

deliver both inpatient and outpatient care spread across multiple wards,

departments and service areas, and they also have more complex management

and leadership arrangements[8]. WASH FIT does not sufficiently consider the

broader health system context and its potential for influencing local change.

In Kenya for example, in larger hospitals, the hospital health management team

comprising the medical superintendent, health administrative officer, nursing

officer in charge and the departmental heads are involved in the day to day

running of the hospital [12]. These teams are assisted by different hospital

committees constituted within the hospitals; these include infection prevention

and control (IPC) committees. The hospital managers and committees prepare

budgets and staffing needs, but the final budgetary and human resource

allocation to these hospitals is the prerogative of regional/county government

[12]. Majority of these larger hospitals in many low- and middle-income

countries have similar organisational arrangements and some similar form of

47

regional administration who have a role in decision making and resource

allocation and need to be involved in the improvement of WASH.

Our report describes an adaptation of WASH-FIT to a Water Sanitation and

Hygiene Facility Survey Tool (WASH FAST). This entails an extension of the tool

to provide a comprehensive assessment of WASH services in hospitals providing

both outpatient and inpatient care. It also provides a mechanism to meet both

local, national and regional needs for tracking WASH improvements. The

adapted tool also considers the complex leadership and management

arrangements. It proposes how responsibilities should be allocated across

different levels of the health system to promote accountability and subsequent

improvement.

Methods

Ethics Statement

For this study, we sought and received informed consent in all cases where this

was relevant. All information received was handled confidentially. All quotes

from the study respondents were anonymised. This study received approval

from the Oxford Tropical research ethics committee ( OXTREC) from the

University of Oxford (Ref: 525-17) and from the Kenyan Medical Research

Institute (Ref: KEMRI/SERU/CGMR-C//086/3450).

48

Adaptation of WASH-FIT into WASH-FAST

The adaptation of WASH-FIT into WASH-FAST entailed: (1) Creating an intuitive

aggregation approach for the WASH indicators, to illustrate variation across

health care facilities and facilitate tracking of WASH over time; (2) Extending

assessment so that indicators are scored for each ward in addition to the facility

as a whole - to highlight potential variation in WASH within a larger facility and;

(3) identifying those responsible for action on WASH with relevant

stakeholders. We illustrate the value of extending WASH-FIT to WASH-FAST by

illustrating how data can be used for identifying challenges and highlighting

variation.

1. Aggregate scoring approach

The WASH-FIT already presents a ‘scoring’ approach with one of three possible

outcomes for each indicator, does not meet target, partially meets target and

fully meets target. The first step involved moving from this qualitative scoring

system to a simple quantitative scoring system that assigns a numeric score to

each indicator based on assessment findings as follows: 0- does not meet the

required standards (i.e. target), 1- partially meets target and 2- fully meets the

target. This enabled us to create aggregate domain scores (based on the

number of indicators within a domain) and aggregate facility scores (based on

all 65 indicators) that can be used to show domain and facilities’ performance.

These aggregate scores can also be colour coded to produce an easy to

interpret “traffic light” reporting approach.

49

2. Identification of ward level indicators

The second step involved identifying which of the existing and 65 WASH

indicators can be assessed at the inpatient-ward level. To select indicators for

assessment in every ward we employed an iterative process to review and

discuss the 65 indicators involving the research team and a team of 19 health

professionals comprising doctors, nurses, pharmacists and public health

officers who had been recruited to pilot test and apply the WASH assessments

in hospitals in Kenya. Using the same simple numeric scoring approach to the

identified indicators as in step 1 above enables aggregate ward scores to be

calculated to help identify variation between wards in the same hospital.

3. Assigning responsibility for action

The third part of the adaptation was to group indicators based on who should

take responsibility for action to improve WASH – addressing the issue of

accountability. For this process, a study team of 4 members familiar with the

Kenyan health care system and its management examined all 65 indicators in a

bid to understand how these indicators relate to one another and assign them

to domains linked to the persons/offices who would be responsible for action

to improve WASH. These levels of responsibility were confirmed through a

series of interviews with health care workers and a subsequent large

stakeholder workshop.

Demonstrating potential and creating tools to help visualise

performance and its variation

We proceeded to collect data using the WASH-FAST tool as part of a survey in

14 county hospitals varying in size and bed capacity across 11 counties during

50

which key informant interviews were also conducted (see below). This survey

is described in more detail in an accompanying paper [13]. In brief, the county

hospitals included are in high and low malaria zones in Kenya (five and nine

sites, respectively). The selection of these hospitals was purposeful and based

on links developed from ongoing work to improve clinical information as part

of a collaboration between the Kenya Medical Research Institute -Wellcome

Trust Research Programme and the Ministry of Health [14]. The survey involved

assembling a team of 7-8 people and conducting a facility assessment at each

hospital. The study team included a leader, four surveyors employed for the

study and 2-3 representatives selected based on their specific role as infection

prevention and control coordinators or public health officers from the

individual hospitals where the survey was being carried out. Data collection

used the same methodology as WASH-FIT and involved direct observation and

discussion with relevant hospital workers to provide clarification of the

assessment where needed. Each indicator was assessed, and the score

determined by team consensus as either not meeting target, partially or fully

meeting the target. Data were collected for each inpatient ward (using 34

WASH indicators), then for indicators assessed at the whole facility level (65

WASH indicators). The 65 facility level indicators included an assessment of

outpatient areas, common service areas (e.g. kitchen, laundry, laboratory,

waste management facilities) and the outdoor environment, taking account of

ward-specific scores where relevant, and represents an overall judgement of

the survey team. The data collection tools and standard operating procedures

used are provided in the supplementary information. (S1 File)

Aggregate scores were generated by summing individual indicator scores and

dividing this total by a denominator that assumed a perfect score for each

indicator.

51

In this way, we then estimate percentage scores for the hospital, WASH domain

and level of accountability using indicators linked to these grouping categories

as appropriate. Summary ward-specific scores were based on individual

indicator assessments made for each ward. The rationale for such sub-scores

was to highlight variation and priority areas for improvement and who should

take responsibility for improvement. To promote the rapid interpretation of

scores, we generated ‘traffic-light’ colour maps presenting percentage scores

using cut-offs of <40%, 40-60%, 60-80% and 80-100%. Data analysis for

visualisation was done using R, an open-source statistical package [15].

Use of qualitative data

Qualitative interviewing pursued two purposes; to understand IPC

arrangements in Kenyan hospitals (with findings reported elsewhere) and to

explore the feasibility and potential value of our proposed allocation of

indicators for accountability.

The interviews were conducted with 17 hospital managers (e.g. medical

directors, nursing and laboratory heads) and 14 frontline health workers (e.g.

consultants, medical and nursing officers) during the survey visits, in seven of

the 16 hospitals – sampled to ensure spread across different geographical

locations represented by the study hospitals.

Interviews were conducted by the first, third, and fifth authors and took

between 30 and 90 minutes. The first author led this section of work and

followed a semi-structured interview approach. The interviews were generally

guided by the ‘long’ or ‘ethnographic’ approach [16], but there was a particular

focus on responsibility for different areas of IPC and WASH, which provided

more structure to the inquiries in this area. All authors have experience with

52

medical research in Kenya, but the first author is a well-experienced doctor with

experience working in different county structures, and he guided the other

interviewers over the course of the interviewing.

The interviews were prefaced by an explanation from a senior member of staff

who had given permission for the research to take place on-site. All

interviewees were then given time to read a background information sheet

concerning the project, and each signed a full written consent. An opportunity

was offered to critique or refuse the interview, or to withdraw permission, but

no respondent chose to do this.

We used both purposive and snowball sampling in order to identify

respondents, and as much as was practically possible, we were mindful of the

mix of gender, age and experience and aimed to reflect this diversity in our

interviewee sampling strategy. Each interviewee was introduced to the

researchers by a senior member of staff familiar with the research, and the

interview took place in or near their place of work. The interviews were

conducted with one or two researchers away from patients and staff. We

ensured in each case that we did not take the respondent away from core tasks

or risk harm to their patients.

No repeat interviews were felt to be necessary, but the interview instrument

was honed to focus on areas of interest over time, allowing us to move beyond

areas where we had reached saturation and onto other new areas. The audio

files were transcribed and uploaded into NVivo 12, and the audio files were kept

on an encrypted laptop. It was relatively trivial to complete our primary goal of

identifying the formal, de jure responsibilities for each level, but it was also

important to code and describe the nuances of the de facto practices that

prevail in the studied sites (as described below).

53

We did not return transcripts to the respondents, but we have sought to share

general findings with hospital management and through the ministry and

county stakeholders with whom we are connected. Most importantly, we used

a stakeholder consultative workshop to confirm or revise the levels of

accountability and related indicator sets. This workshop included

approximately 120 technical experts and key stakeholders in WASH comprising;

Ministry of Health officials, Hospital WASH leaders, county health department

leaders, and doctors and nurses in Kenya with interest in IPC. This cross-

checking of indicator allocation to different levels of responsibility was

completed before creating scores for these domains. We also used the

stakeholder workshop to get feedback on the use of aggregate scores and data

visualisation approaches, confirming that the proposed reporting methods

would be of value to potential end-users.

The interview guides used for the study are available as a supplement. (S2File)

A (Consolidated criteria for Reporting Qualitative research) COREQ checklist

was successfully completed and is included as an appendix. (S3 File)

Results

From the tool redesign to collect data at ward level, we established that 34 of

the 65 indicators could also be assessed at the ward level. A description of all

65 indicators is provided as a supplement. Table 1 below provides a summary

of the number of indicators that were to be assessed at the ward and facility

level by the original WASH-FIT domains and by the proposed levels of

responsibility.

54

Responsibility for action

We developed a reorganisation of the existing WASH indicators based on their

logical relationship and who would be responsible for action resulting in a

classification with three levels of responsibility. These are, first, the county

government which should be concerned with indicators that are beyond the

control of hospital leadership (this level might be a national government where

resources are not fully devolved). The second level is the hospital health

management team (the medical superintendent, health administrative officer,

the nursing officer in charge and the departmental heads) and the last level is

the hospital infection prevention and control committee (Table 1).

On the proposed levels of responsibility in the WASH-FAST, although 2 of the 9

indicators under the responsibility of the county government could also be

assessed at ward level, these are (i) water services available in sufficient

amounts and (ii) having rewards for high performing staff, these only need to

be assessed at the facility level for tracking progress in follow up assessments.

Therefore, when grouped by the responsibility, we suggest only 32 of the

original 34 indicators are assessed at ward level (Table 1).

55

Table 1: Summary Indicators at ward and facility level by WASH domains and WASH-FAST

WASH-FIT WASH-FAST

WASH DOMAINS WASH DOMAINS ACCOUNTABILITY DOMAINS

Facility Ward Facility Ward Facility

Water 14 Water 6 14 County

Government

0 9

Sanitation &

Health care

Waste

22 Sanitation &

Health care

Waste

11 22 Hospital

Management

16 31

Hand hygiene,

Environmental

Management,

Cleanliness

and

Disinfection

18 Hand hygiene,

Environmental

Management,

Cleanliness

and

Disinfection

12 18 Infection

prevention &

control

committee

16 25

Organisational

Management

11 Organisational

Management

5 11

Total 65 Total 34 65 Total 32 65

The in-depth interviews allowed us to explore the relationships between the

WASH criteria and to establish where responsibilities lay for each. This

contributed to the emerging model of the layers of WASH management and

informed our understanding of the causalities and contingencies in this area.

County Level – The County is responsible for setting the budget for each

hospital, and importantly, sets the overall budget for health spending. This

impacts on general, but hugely important, WASH-related criteria such as

56

staffing levels and material upkeep of hospitals. Additionally, while each

department in each hospital is asked to project their needs for the next year as

part of hospital budgeting processes, the requested amounts may be ignored

by counties. Hospitals thus needed to work within the limitations of the budget

and staffing allowed them.

“You know normally we are told to itemise whatever we

require in the departments that we are working in …yes, by

different departments, come up with their budget proposal.

The administrator compiles the budget for the whole

hospital and then give it…we don’t control funds in the

institution. Every finance that is channelled to the hospital is

controlled by the chief officer in the county. So, we send the

budget to the county” Hospital Manager

Whereas the day to day running of the hospital is done by the hospital

management, some of the activities are delegated to committees within the

hospital.

Hospital Level – Key areas of hospital management were in part delegated to

committees that held responsibilities for activities and addressing needs. The

effectiveness of IPC committees in different hospitals appeared to be variable,

but where they were operational, they had an essential influence on resource

allocation and monitoring of WASH.

“…when the committee, the IPC committee, meets they raise

their needs as per various departments, and then the hospital

now addresses that. Like if you want to purchase, for instance,

you want bins, litter bins, disposal bags, waste disposal bags.

57

So, you raise your needs as per your department because you

know different departments have got their different needs”

Hospital Manager

However, despite their importance, these IPC committees in some facilities

struggled to gain respect relative to other more prestigious committees and

were regarded to be of low status.

There are some committees which are found to be more,

which are more do I say prestigious? They look better. So, if

I am in IPC, people will be thinking okay… so IPC will have no

one. I mean, what is the benefit of being in IPC, what is there,

how am I gaining being in IPC? Consultant

Ward Level – Interest and capability at the ward level is essential to effective

WASH. The individuals responsible for WASH at this level are not likely to have

the ability to affect budgets and resource allocation, but they are essential in

both maintaining supplies and overseeing important areas such as hand

hygiene. Variability in performance at ward level may be linked to the presence

of an individual in the ward who has interest and passion for IPC related

activities.

“And we also have someone, he’s also a team leader in the

infection control and making sure we have... whenever he’s

available we have our sanitisers, make sure we have soap,

make sure we have gloves” Frontline Health Worker

The relative importance of IPC varied ward to ward; however, with the newborn

units (NBU) often used as an example to contrast high versus low performance:

58

“Across the hospital, in NBU is where I know there is strict

infection prevention because once you are getting into NBU,

you remove your lab coat, you wash your hands and then you

get into the unit where you fold whatever you are wearing, a

long-sleeved anything you fold it, and then you get in the

unit… Now in other wards, we don’t have such strict infection

prevention, you get in, and you start...” Frontline Health

Worker

Consultative Workshop

The consultative workshop was held in November 2018 during the annual

national IPC symposium. There were 120 people in attendance. These included

Ministry of Health officials, managers from the hospitals and county

government, development partners and training institutions who are familiar

with infection prevention and control and WASH and frontline health workers.

The workshop attendees discussed, amended and approved the proposed

levels of accountability (Table 1) and made specific recommendations that

hospitals identify a champion to lead the IPC committees and to identify ways

of boosting morale for IPC related issues among health workers across these

hospitals.

Based on all of this work, a final indicator framework was developed (Fig 2)

that shows the relationship between the indicators, their original WASH-FIT

domains, and how they are allocated to different levels of responsibility. We

also use Fig 2 to highlight which of the individual indicators can also be

assessed at ward level.

59

Fig 2: Schematic layout of WASH FIT indicators. Illustrates how indicators assessed at ward and facility level are logically related. These are grouped by the original four domains and by levels of responsibility. The indicators with a red bold outline were also assessed at ward level. The dotted boxes are used to describe categories and are not part of the indicators.

60

Visualisation approaches to support monitoring

Using an example of data collected from four of the 14 hospitals, two large (H2,

H9) and two small (H1, H7) hospitals, we present an illustration (Fig 3) of how

performance of two domains (water and sanitation) vary between hospitals

(Panel A) and how the individual wards within these facilities performed (Panel

B). We note differences between domains and differences between hospitals,

with some facilities having scores of <50%. We also note variability across wards

in these hospitals. From this example (Panel B) for the water domain ward

scores in hospital H1 show, minimal variability compared to those of hospital

H9. We contrast our visualisations with data presented using the original

WASH-FIT template at the facility level for the four hospitals in Fig 3, Panel C.

61

Fig 3 Service performance variation by ward and hospital and the original WASH FIT scores. Panel

A: Radar Plot of facility-level scores from four hospitals for two domains (Water, Sanitation)

showing similar performance for sanitation overall but more marked variation for water varying

hospital performance. Panel B: Shows ward domain scores from multiple wards (dots) for two

domains (water, sanitation) illustrating their variation, the mean of these ward-specific scores (

circled cross) and the overall facility aggregate score (plain circle O). The overall facility score

includes an assessment of inpatient wards and other service areas (kitchen, outpatient, outdoor

environment) across the hospital. Panel C shows WASH-FIT facility-level scores of four hospitals

for the two domains.

To further illustrate how WASH-FAST can provide detailed information for use

at national and regional levels on hospital performance and where

62

responsibility for action lies, we present an example of all the 16 indicators

[spanning all the WASH domains] under the IPC committee at ward level. Here

we generate the summary ward scores for each of the four hospitals (H1, H2,

H7, H9) coded using a traffic light colour system with red being a score of <40%

and green indicating a score of >80%. Fig 4 illustrates a ‘dashboard’ approach

that shows performance across the hospitals for the individual IPC-committee

related indicators assessed at ward level (horizontal bar chart, for example

highlighting a need for the IPC committees to work on availing cleaning records

in the wards in these four hospitals). It also helps visualise the overall mean

ward scores for each hospital for all 16 indicators (the top panel vertical bar

chart). The central traffic light coding presents a summary of how the individual

indicators performed in each specific hospital.

63

Fig 4. Performance of IPC domain indicators. Ward level indicator performance (upper bars). The right bars summarise indicator performance across hospitals. The central grid shows individual indicator performance. SOP: standard operating procedures, ITN: insecticide-treated nets

64

Discussion

We have presented an adaptation of the Water and Sanitation for Health

Facility Improvement Tool (WASH-FIT) into WASH-FAST (Facility Survey Tool).

The adaptation entailed an extension of the tool to meet primarily national (i.e.

situation analysis, monitoring and tracking) needs, and to facilitate

comprehensive assessment of WASH services in larger – more complex –

secondary and tertiary health care facilities s encompassing both outpatient

and inpatient care and multiple medical specialties. The adapted tool scores

indicators at various levels of the facility (including by ward and by medical

specialty) and assigns levels of accountability for each indicator, to identify

what services can be addressed by whom locally or at higher levels of the health

system. An aggregated numeric scoring system, consisting of a percentage

score out of the total that would be obtained if all indicators met the expected

target, can be used to identify service areas requiring priority action within a

facility or to identify facilities or specialties requiring priority action nationally

or sub-nationally.

Adequate governance and leadership are one of the foundations for the

provision of quality care. Governance for quality includes improving

accountability and identifying the roles and responsibilities at all levels of health

systems and using data to make decisions [17]. Thus WASH-FAST may also be

used to identify responsible actors limiting or effecting positive change within

a facility and beyond, and to potentially reward excellent performance in a bid

to improve staff morale concerning IPC/WASH. WASH-FAST assumes that

performance and quality indicators for WASH are the responsibility of three

possible actors. These are, administrative division officers or governments

responsible for budgetary and human resource allocation to hospitals, senior

65

hospital management teams and relevant facility-based specialised committees

or groups of persons who are essential in decision making for IPC related

activities, such as the infection prevention and control committee. Although

the WASH-FAST was developed within a Kenyan context, we expect these broad

accountability domains (endorsed by government representatives, public

health officers, IPC experts and health care professionals through interviews

and a consultative workshop), to be generally applicable to most low- and

middle-income countries, with minor context-appropriate considerations. This

would allow more comprehensive use of WASH-FAST and could support within

or between-country comparisons where relevant.

We anticipate that aggregated scores derived from the application of WASH-

FAST can be used more broadly to inform health system leaders on whether

and what facilities and specialties require action at either local, sub-national or

national level to improve WASH services. The simple scores allow the

comparison of WASH services within and between facilities and or medical

specialties either cross-sectionally or over time (i.e. to identify changes in

quality and performance and trends), through repeated surveys. The extended

tool is hence a broadly applicable facility improvement tool – potentially

encompassing training, team building and risk assessment steps as per the

WASH-FIT process - that also appertains to WASH performance monitoring sub-

nationally and nationally. Training of health care facilities staff to partake in

surveys and facility improvement plans, in turn, empowers and encourages

staff to take interest and ownership on WASH and IPC, contributes to up-skilling

in these areas and improves short and long term sustainability of interventions

and developments where applicable [18]. WASH-FAST may also be applied to

help remedy the paucity of data on the status of WASH services in low- and

middle-income countries, help bridge evidence-based gaps and provide a

66

platform to monitor interventions aimed at improving AMR and patient safety.

This is while continuing to serve the original purpose of continuously informing

a local improvement plan in small primary health care facilities as well as more

extensive facilities comprising multiple wards and medical departments.

A limitation of both WASH-FAST (and WASH-FIT), is that the score assigned to

selected individual indicators may be subjective, where it relies on observations

that could vary from person to person. To mitigate this, we developed standard

operating procedures before data collection, conducted training for the data

collection teams and used consensus among surveyors to assign scores during

the assessments. WASH-FAST also rests on the premise that the hospitals have

well-structured leadership, including a functioning infection prevention and

control committee or relevant expert group. The indicators are also not

weighted in accordance to the health risk they pose, implying that identical

aggregate scores may have very different decision-making implications

depending on the composite of indicators considered in the score (e.g.

availability of water vs cleaning protocols). The same limitation applies to

repeated measurement, where a facility may get the same score over two

consecutive surveys, perhaps reflecting improvements in some areas but

worsening in others. To mitigate these limitations, aggregated (summary)

scores should be interpreted in the context of individual indicator scorings

presented through heatmaps or other visualisation tools.

We appreciate that although we adapted an already validated tool (WASH-FIT),

there is need to extend the use of the WASH-FAST to other populations and

settings to apply the tool under operational conditions, not by researchers.

Although it was not our aim to validate the WASH-FAST, we explored key

elements of face validity as described by Nevo [19]. We involved stakeholders,

experts and health workers in the consultative workshop, to check if the items

67

in the WASH-FAST were appropriate (rater involvement). The research team

and the stakeholders also assessed if the content in the tool, including the levels

of responsibility, was valid (hypothesised validity)[19]. The other major part of

face validity involves establishing if the method of measurement is appropriate,

and for this, we established through the consensus in the workshop that the

WASH-FAST was indeed suitable for assessing WASH practices. However,

looking at the nature of the tool and similar tools, where there are no gold

standard measures, validation may be challenging. It may need to be sufficient

that there is a consensus on the value of addressing the issues identified by

indicators. This may be thought of more as similar to an appraisal than a true

measurement of WASH performance on a linear scale.

To improve hospitals as platforms that provide high-quality care and prevent

the emergence of AMR, proper WASH, and IPC structures are core [17]. We

suggest that using WASH-FAST to monitor and improve the capacity for WASH

and IPC would enhance governance for quality and limit the emergence of AMR

by promoting accountability and identifying the roles and responsibilities at all

levels of the health systems [17]. In the process of accelerating universal health

coverage in many counties, hospital accreditation has become a key component

as it provides for insurers and governments a criterion for which hospitals to

include in their funding mechanisms. WASH-FAST can thus be used as part of

the tools for hospital accreditation to ensure they focus adequately on IPC

structures as part of preventing patient harm and AMR [20] [21].

Conclusion

We propose the WASH-FAST (Survey) tool as a modification/extension of the

original tool. Compared to the WASH-FIT, WASH-FAST provides for additional

assessment of WASH within the hospital and assigns responsibility for action.

68

Its use is most relevant in larger hospitals that have multiple inpatient

units(wards) as it allows for assessment at the ward level in addition to the

overall facility assessment. Due to its ability to provide aggregate scores, it can

be used to monitor and track the progress of WASH at the hospital, regional or

national levels providing crucial data for governments and international

development agencies who provide support for WASH. In addition, due to the

ability to assign responsibility for action, WASH-FAST allows for persons/teams

to take responsibility in improving the state of WASH at the hospital and

regional level. Where the primary aim is to support local improvement in

smaller facilities, WASH-FIT remains the tool of choice.

Acknowledgements

We thank the Ministry of Health and the council of governors who permitted

this work to be carried out. We also thank the hospital management and clinical

teams who supported the work in the survey hospitals. This work is published

with the permission of the Director of KEMRI.

Author contributions

The roles of the contributors were as follows: M Maina, J.M, O.T, C.S and M.E

conceived the study. M. Maina, G.K and M.Z, collected data, A.H and M

Montgomery, provided technical advice on the methodology, O.T, J.M assisted

M.M in data analysis and interpretation. M. Maina drafted the manuscript. M.

Maina, J.M, O.T, C.S, A.H, M. Montgomery and M.E critically revised the

manuscript for intellectual content. All authors read and approved the final

manuscript.

69

Data Availability

All summary data underlying the findings are freely available in the

manuscript and supplemental files. The raw data used for this manuscript are

hosted in a public repository Harvard Data verse. DOI Information:

https://doi.org/10.7910/DVN/IJUWWR

70

References

1. United Nations General Assembly. Transforming our world: the 2030 Agenda for

Sustainable Development. New York: United Nations. 2015;(1).

2. Hans Jørn Kolmos. Health Care Associated Infections: Sources and Routes of

Transmission. In: Infection Control – Updates Croatia2012 [cited 2017 2nd November ]. Available

from: http://cdn.intechopen.com/pdfs-wm/28876.pdf.

3. Rainey R and Weinger M. The Role of Water, Sanitation and Hygiene (WASH) in

Healthcare Settings to Reduce Transmission of Antimicrobial Resistance. 2016 [cited 2017 18th

December]. Available from: http://resistancecontrol.info/infection-prevention-and-control/the-

role-of-water-sanitation-and-hygiene-wash-in-healthcare-settings-to-reduce-transmission-of-

antimicrobial-resistance/.

4. WHO/UNICEF. Water, Sanitation and Hygiene (WASH) in Health Care Facilities Global

Action Plan [cited 2018 28th September]. Available from:

http://www.who.int/water_sanitation_health/healthcare_waste/wash-in-healthcare-facilities-

action-plan.pdf.

5. Bouzid M, Cumming O, Hunter PR. What is the impact of water sanitation and hygiene

in healthcare facilities on care seeking behaviour and patient satisfaction? A systematic review of

the evidence from low-income and middle-income countries. BMJ Global Health. 2018;3(3). doi:

10.1136/bmjgh-2017-000648. PubMed PMID: 29765776; PubMed Central PMCID:

PMCPMC5950627.

6. World Health Organization. WASH in health care facilities: Links with the network for

improving quality of care for maternal newborn and child health 2017.

7. World Health Organization, UNICEF. Water, sanitation and hygiene in health care

facilities: status in low and middle income countries and way forward. 2015.

8. World Health Organization, UNICEF. Water and Sanitation for Health Facility

Improvement Tool (WASH FIT): a practical guide for improving quality of care through water,

sanitation and hygiene in health care facilities. Geneva: 2017 9241511699.

9. Weber N, Patrick M, Hayter A, Martinsen AL, Gelting R. A conceptual evaluation

framework for the water and sanitation for health facility improvement tool (WASH FIT). Journal

of Water, Sanitation and Hygiene for Development. 2019. doi: 10.2166/washdev.2019.090.

10. John A, Jamie B, Yves C. Essential environmental health standards in health care.

Geneva: World Health Organization. 2008.

11. World Health Organization. Guidelines on core components of infection prevention and

control programmes at the national and acute health care facility level: World Health

Organization; 2016.

71

12. Barasa EW, Manyara AM, Molyneux S, Tsofa B. Recentralization within

decentralisation: County hospital autonomy under devolution in Kenya. PLOS ONE.

2017;12(8):e0182440. doi: 10.1371/journal.pone.0182440.

13. Maina M, McKnight J, Tosas-Auguet O, Zosi M, Kimemia G, Mwaniki P, et al. Evaluating

the foundations that help avert antimicrobial resistance: Performance of essential water

sanitation and hygiene functions in hospitals and requirements for action in Kenya

2019:Forthcoming.

14. Tuti T, Bitok M, Malla L, Paton C, Muinga N, Gathara D, et al. Improving documentation

of clinical care within a clinical information network: an essential initial step in efforts to

understand and improve care in Kenyan hospitals. BMJ Global Health. 2016;1(1). doi:

10.1136/bmjgh-2016-000028.

15. R Core Team. R: A language and environment for statistical computing. R Foundation

for Statistical Computing, Vienna, Austria. 2013.

16. McCracken G. The Long Interview.: Sage; 1988.

17. Kruk ME, Gage AD, Arsenault C, Jordan K, Leslie HH, Roder-DeWan S, et al. High-quality

health systems in the Sustainable Development Goals era: time for a revolution. The Lancet

Global Health. 2018;6(11):e1196-e252. doi: 10.1016/S2214-109X(18)30386-3.

18. de Francisco Shapovalova N, Meguid T, Campbell J. Health-care workers as agents of

sustainable development. The Lancet Global Health. 2015;3(5):e249-e50. doi: 10.1016/S2214-

109X(15)70104-X.

19. Nevo B. Face Validity Revisited. Journal of Educational Measurement. 1985;22(4):287-

93.

20. Smits H, Supachutikul A, Mate KS. Hospital accreditation: lessons from low- and middle-

income countries. Globalisation and Health. 2014;10(1):65. doi: 10.1186/s12992-014-0065-9.

21. National Health Insurance Fund. Assessment checklist for accreditation of Health

Facilities ( NHIF ACT 1998). Nairobi: NHIF; 2016.

72

Interviewee code Institution Position /role

Grade (where applicable) Age

Gender

Highest education / training

Years of experience in Infection Prevention and

Control

Membership to relevant organisations

Core competencies in infection prevention and control

(Tick the option that best defines the interviewee’s role)

Management / Supervision

Frontline contact, service or care

Comments:

Supporting Information S2

Semi-Structured Interview Guide for Hospital Managers and Frontline Healthcare Workers

C losed Questions

73

1. What IPC guidelines are available in this hospital? In your view, what guidelines/documents do you feel have affected the way you practice in this hospital?

Step through all general and domain-specific policy documentation and enquire about potential additional sources of information. >>

2. Do you have an IPC team? When did they last meet?

Ask who are members, when it meets, what roles and responsibilities are? >>

3. How are doctors, nurses and other healthcare workers trained for IPC at this hospital? How are they trained on the new guidelines? As a student?

Ask what percentage of staff have received; if renewed; who gave the training.>>

4. Is there a specific IPC budget? How much is it annually?

Investigate annual budget setting, how funds are allocated in general. Is IPC only funded as a management area by external sources? >>

5. What are the main IPC problems you experience at the hospital level?

<<Step into each one. Ask for examples, details on what the problems are and how they’re dealt with>>

Probes

Following the available guidelines by the clinicians? Funding

Lack of supply or poor-quality Equipment for IPC activities, e.g. gloves, masks

6. What do you think patients know about IPC?

<<Exploratory – looking for background. If they use phrases like ‘ignorant’, ask what they mean by this. If they differentiate between themselves and the patients, ask why nurses/doctors are different.>>

74

<<Explain AMR – infections becoming resistant to drugs – you know about Multi-Drug Resistant TB but happening to a lot of infections. Very difficult to treat – running out of antibiotics. IPC is one way of dealing with AMR, but we’re also interested in how you use antibiotics…

7. Have you heard of antibiotic stewardship? (yes – move to 8; no – explain ABS is where we’re careful we don’t use antibiotics for infections that don’t need them, or use the wrong antibiotics or don’t make sure patients finish their course).

8. Are there any guidelines or training for ABS (correct and rational use of antibiotics)? Is this dealt with as part of IPC?

By training, this may include CME, seminars and external training

<<Explore any differences. Be prepared to explain what ABS is. >>

9. How are doctors, nurses and other healthcare workers trained for ABS at this hospital? How are they trained on the new guidelines? As a student?

Ask what percentage of staff have received; if renewed; who gave the training.>>

10. Who sets policy for ABS in the hospital? What guidelines, if any are, they currently using for prescription of the antibiotics? How are you able to identify if antibiotics are not working? How is antibiotic resistance dealt with in this hospital?

Identify documents and guidelines in actual use first. Then ask if they are aware of national guidelines etc.

11. What do you think patients should know about antibiotics? Do you have patients that have preferred antibiotics? Do they know it’s important they should finish the course of medication? Do they sometimes stop and start? How do you get them to finish the course of medication?

75

<<As above, focus on any accusations of ‘ignorance’ or ‘lack of education’. Ask what this means, and how doctors/nurses deal with this. >>

12. I can imagine the workers at this hospital are very busy and have lots on their minds. How much do you think that IPC and ABS are prioritised at this hospital? Do medical staff see IPC as something important? how? why?

<<Do medical staff take responsibility for IPC? >>

(Thank the interviewee for their time, explain the project again, and inform them that they will be made aware of findings.)

76

77

Chapter 3

Evaluating the foundations that help avert

antimicrobial resistance: Performance of essential

water sanitation and hygiene functions in hospitals

and requirements for action in Kenya

Michuki Maina, Olga Tosas-Auguet, Jacob McKnight, Mathias Zosi,

Grace Kimemia, Paul Mwaniki, Constance Schultsz and Mike English

PLoS One. 2019;14(10): e0222922.

https://doi.org/10.1371/journal.pone.0222922

78

Abstract

Background

Water Sanitation and Hygiene (WASH) in healthcare facilities is critical in the

provision of safe and quality care. Poor WASH increases hospital-associated

infections and contributes to the rise of antimicrobial resistance (AMR). It is

therefore essential for governments and hospital managers to know the state

of WASH in these facilities to set priorities and allocate resources.

Methods

Using a recently developed survey tool and scoring approach, we assessed

WASH across four domains in 14 public hospitals in Kenya (65 indicators) with

specific assessments of individual wards (34 indicators). Aggregate scores were

generated for whole facilities and individual wards and used to illustrate

performance variation and link findings to specific levels of health system

accountability. To help interpret and contextualise these scores, we used data

from key informant interviews with hospital managers and health workers.

Results

Aggregate hospital performance ranged between 47 and 71% with five of the

14 hospitals scoring below 60%. A total of 116 wards were assessed within

these facilities. Linked to specific domains, ward scores varied within and across

hospitals and ranged between 20% and 80%. At ward level, some critical

indicators, which affect AMR like proper waste segregation and hand hygiene

compliance activities had pooled aggregate scores of 45 and 35% respectively.

From 31 interviews conducted, the main themes that explained this

heterogenous performance across facilities and wards included differences in

79

the built environment, resource availability, leadership and the degree to which

local managers used innovative approaches to cope with shortages.

Conclusion

Significant differences and challenges exist in the state of WASH within and

across hospitals. Whereas the senior hospital management can make some

improvements, input and support from the national and regional governments

are essential to improve WASH as a basic foundation for averting nosocomial

infections and the spread of AMR as part of safe, quality hospital care in Kenya.

Keywords

Water Sanitation and Hygiene (WASH), Infection prevention and control,

Antimicrobial Resistance

80

Introduction

Water Sanitation and Hygiene (WASH) services in healthcare facilities are

integral in the provision of safe, high-quality healthcare and an essential

foundation for averting the spread of antimicrobial resistance (AMR). Facilities

with inadequate WASH are associated with a higher risk of hospital-associated

infections and increased environmental contamination from clinical waste [1,

2] [3]. A recent global report was published by the World Health Organization’s

Joint Monitoring Programme for Water Supply, Sanitation and Hygiene (JMP)

highlighting the current challenges with WASH. It reviewed data from 56,000

health facilities in 120 countries and revealed that a quarter of health facilities

assessed lacked water from an improved source on the premises and almost

half lacked hand hygiene facilities at the points where care is provided [4]. The

bulk of these WASH challenges reported are in Asia and Sub Saharan Africa [4].

To improve the state of WASH in health facilities, the World Health

Organization(WHO) and United Nations Education Fund (UNICEF) developed

the Water Sanitation and Hygiene Facility Improvement Tool (WASH-FIT) [5].

The tool was based on global standards of environmental health and infection

prevention and control (IPC) [6, 7]. It was designed for primary care facilities in

limited-resource regions to assess the state of WASH and promote self-

improvement. The process of developing the WASH-FIT was an iterative

process involving different stakeholders to ensure the tool was validated before

rolling out.

WASH-FIT entails a process of self-assessment that focuses on achieving

minimum standards for a clean and safe environment in primary care facilities.

The tool was not designed for more extensive facilities with multiple inpatient

units, and it was not meant to survey and compare WASH performance across

81

hospitals and their departments. We, therefore, modified the tool to create a

WASH facility survey tool (WASH FAST) for use in surveys in larger hospitals

with multiple inpatient units; collecting data both at the ward and facility level

and creating levels of responsibility to improve accountability for WASH [8]. The

development of the WASH-FAST tool has been described in detail elsewhere

[8]. Briefly, this entailed three main steps: The first step was developing an

approach to produce aggregate numeric scores, to enable comparisons and

tracking of hospital performance of WASH over time. Secondly, it involved

modifying the assessment of hospitals providing multi-speciality care - so that

relevant indicators are assessed and scored for each ward in addition to the

facility as a whole. Finally, the adapted tool identifies the actors who are

accountable for the issues uncovered and so are responsible for effecting

positive change in WASH. A comparison of the main differences between the

WASH-FIT and WASH-FAST is illustrated in Fig 1.

Fig 1. Comparing the WASH-FIT and WASH-FAST tools

We herein present the findings of a survey that investigated the provision of

water, sanitation, hygiene and their management in 14 Kenyan county-level

hospitals using WASH-FAST. For brevity, we present a subset of the results to

illustrate the application of WASH-FAST and to describe the status and

82

consistency of WASH services within and across hospitals. Full results at ward

level are provided in as a supplement (S1 Table). We draw on in-depth

interviews with hospital managers and frontline healthcare workers conducted

in parallel to the survey, to help interpret and contextualise WASH-FAST results.

We also show how the findings on WASH performance can be linked to different

levels of accountability within the health system. Finally, we identify best and

worst-performing indicators across the hospitals based on the scoring system,

some of which are critical to safe and quality patient care with implications for

emergence and transmission of antimicrobial resistance (AMR). The

quantitative and qualitative results combined provide important insights on

WASH to national and regional governments and hospital managers. Findings

can be used to inform prioritisation of actions and resource allocation aimed at

improving patient safety and reducing hospital-associated infections and AMR.

Methods

Ethics Statement

For this study, we made every effort to ensure the quality and integrity of the

research. We sought and received informed consent in all cases where this was

relevant. We respected the confidentiality and anonymity of our research

respondents and checked they were willing to participate in the study

voluntarily. We made every effort to anonymise quotes from the study

respondents. This study received approval from the Oxford Tropical research

ethics committee ( OXTREC) from the University of Oxford (Ref: 525-17) and the

Kenyan Medical Research Institute (Ref: KEMRI/SERU/CGMR-C//086/3450).

83

Study Setting

The survey was carried out across 14 public county hospitals (formerly district

hospitals) in Kenya. Hospitals are located in high and low malaria-endemic

regions and represent a diverse selection with varying bed capacities both in

urban and rural areas. All facilities participating in the study provide multi-

speciality care, including at least maternity services and inpatient neonatal,

paediatric, medical and surgical care. The hospitals are part of the Clinical

Information Network (CIN) of the Kenya Ministry of Health. The CIN was set up

to collate data from paediatric inpatient units to promote development and

adoption of evidence-based clinical guidelines [9] and is coordinated by the

Kenya Medical Research Institute (KEMRI) Wellcome Trust Research

Programme.

Survey Preparation and Data Collection

As part of WASH-FAST, a data collection tool and instructions booklet were

designed to allow the systematic and reproducible collection of data across

survey hospitals. The booklet outlines the steps that are to be followed by the

survey team, from the time they reach the facility, to undertake a systematic

assessment of WASH in all hospital areas (S2 File). The WASH-FAST booklet

contains 34 indicators for each hospital inpatient ward grouped into four WASH

domains, to be assessed and scored in agreement to written guidance (water

[6]; sanitation [11]; hygiene [12]; organisational management [5]). An overall

facility-level assessment guide at the end of the booklet was must also be

completed, which contains 65 indicators also grouped into the 4 WASH domains

(water [14]; sanitation [22]; hygiene [18]; organisational management [11]).

The facility-level scoring entails a global assessment spanning all hospital areas,

including outpatient services, kitchen, laundry, waste disposal infrastructure,

84

the outdoor environment, and prior assessment of inpatient wards. Written

guidance helps surveyors allocate a +, ++ or +++ score to each indicator, which

corresponds to 0, 1 or 2 in the numeric scale used to calculate aggregate and

percentage scores. The survey booklet provides text boxes to document field

findings and explain the rationale for individual indicator scores. Indicator

scores can also be used to assess specific WASH domains and sets of indicators

linked to different levels of accountability. At ward level, 32 of the 34 ward

indicators can be assigned for action/accountability. We use 16 indicators

linked to the overall hospital management domain and 16 to the infection

prevention and control committee (IPC) to explicate these uses [8].

A total of 19 health workers comprising doctors, nurses, pharmacists and public

health officers, were recruited before the study to assist with the survey. The

health workers were seconded to take part from the participating hospitals

based on their familiarity with, and interest in improving infection prevention

and control and WASH in their premises. Prospective surveyors were trained by

research team supervisors in February 2018 for one week. The training

comprised a theoretical and practical introduction to WASH, the WASH-FAST

tool and the survey standard operating procedures and included participation

in a one-day survey pilot at a district hospital with similar traits to facilities

participating in the study. WASH-FIT training modules from the World Health

Organisation (WHO) were adapted for the training of surveyors, while research

team supervisors received WASH-FIT training directly from WHO.

Data collection commenced one week after the training and pilot and continued

over two months (February- March 2018). Due to practical and logistical

considerations, surveyors were divided into three teams, each composed of 4-

5 surveyors plus one research team supervisor. Teams were then allocated a

sample of hospitals for assessment at either western Kenya (five Hospitals),

85

central Kenya (four hospitals) or around the capital city Nairobi (five Hospitals).

At survey sites, the surveyor team was joined by 2-3 hospital representatives

with specific roles as infection prevention and control coordinators or public

health officials. This approach facilitated the training of local focal persons and

built on-site capacity to undertake to follow up WASH assessments

independently in future.

Data collection began with a meeting with the hospital management to collate

information on the layout of the facility and retrieve the complete list of wards

and service areas. This was followed by a walk-through of the hospital, noting

the general external and indoor environments as well as any new or old

buildings and infrastructure. A thorough assessment of each eligible ward was

then conducted. The assessment included inpatient wards in the paediatric,

medical, surgical and neonatal departments but excluded units not present in

all hospitals (i.e. critical care, Ear Nose and Throat (ENT), eye, renal and

psychiatric units). In each eligible ward, ward assessment forms were

completed. Once these ward level inspections were complete, there was an

inspection of the entire facility, including the laundry, kitchen, outpatient areas

and the external environment. Each indicator was assessed by direct

observation and the score determined by team consensus on a three-point

scale (meets = 2, partially meets = 1, or does not meet = 0 the required

standard). A detailed explanation of the rating given was also provided in text

notes by the data collection team. In each facility, data were collected over four

days.

WASH is highly dependent on a range of health systems factors [10]. To

understand the causes for the performance outcomes we measured, and the

context of the survey results, the underlying health system components that

support WASH activities were also investigated, including the availability of

86

‘soft’ and ‘hard’ infrastructure, material resources, local guidelines, and

appropriate budgets. In addition, interviews were conducted with 17 hospital

managers (e.g. medical directors, nursing and laboratory heads) and 14

frontline health workers (e.g. consultants, medical and nursing officers) in

seven of the 14 hospitals. The sample was chosen to capture different socio-

demographic influences on WASH and to determine the generalisability of the

observations across the survey sites. To this end, interviewees also comprised

a balanced mix of gender, age and experience. The design of the original semi-

structured interview instrument was informed using informal discussions with

stakeholders and experienced Kenyan clinicians within the research team. The

instrument was subsequently revised organically throughout the research to

reflect new insights garnered through earlier interviews. These interviews were

conducted by the first, third and fifth authors. Each interview lasted 30-90

minutes. All the authors are well vast with the Kenyan context and have

conducted research in Kenya. The first author is a Kenya doctor with vast

experience working in the Kenyan health system and offered guidance to the

interviewers during this process. All the interviewees were approached under

the guidance of the hospital directors. Background information concerning the

study and interview was availed to these interviewees before consenting, which

was done in writing. None of the respondents declined to give informed

consent. These interviews were conducted in a quiet area of the hospital by one

or two members of the study team and were audio-recorded.

There were no repeat interviews conducted during this study after reaching

saturation. Although the transcripts were not returned to the respondents,

general anonymised feedback of the study was provided to hospital

management in each hospital.

87

All interview tools and information sheets are available on request. A

(Consolidated criteria for Reporting Qualitative research) COREQ checklist was

completed and is included as a supplement. (S3 File)

Data Quality Assurance

During the study, data on WASH were entered into paper forms in the booklets

provided. At the end of each day, the study coordinators together with the

teams reviewed the data entered to ensure all the indicators were assessed and

correctly documented. In case of any missing or unclear entries, the study team

made specific reassessments the following day to obtain the data. This was

done for all sites before moving to the next hospital. After completion of data

collection, data from the paper forms were entered electronically into a

database. To ensure data quality, double entry was done and counterchecked

by two members of the study team. All the interview recordings were stored

securely before being transcribed verbatim.

Data Analysis

For analysis, to compute aggregate scores, the three indicator levels: Does not

meet the target, partially meets and meets target were assigned numeric scores

0, 1, and 2 respectively and individual indicator scores summed within the pre-

specified levels of whole facility or ward or WASH domains or levels of

responsibility. There were two main scoring approaches used for data

analysis.[8]

1. Aggregate percentage scores within hospitals generated at

ward and facility level.

Percentage scores were derived after summing the numerators and

dividing these by the sum of denominators (representing the maximum

88

possible score) for sets of individual indicators as described above. We

used four categories to generate a colour chart with a “modified traffic

light” system to display results using cut-offs of <40%-red, 41-60%-

yellow, 61-80%-orange and 81-100%-green.

2. Aggregate performance across hospitals at domain and

indicator level.

The overall domain percentage across the 14 hospitals was derived

after summing the individual hospital indicator scores for each of the

four domains (numerator) presented as a proportion of the maximum

possible score. At the indicator level, the overall indicator percentage

by domain was computed as the total score of each indicator from all

the hospital wards as a proportion of the total maximum possible score.

These results are presented using simple histograms/bar charts to represent

results and variations, and in the case of ward level percentage score, a scatter

plot was used to describe each score and their variation and median score. All

quantitative data analysis was conducted in R [11].

Qualitative Data

The audio files were transcribed and uploaded into NVivo 12 [12], and the audio

files were kept on an encrypted laptop. The third and fifth researchers coded

the transcripts independently before discussing the codes and agreeing on

combined axial codes. These discussions also involved the first author who used

his knowledge of the survey data to make links between the survey findings and

the open codes generated from the transcripts. We then arranged the axial

codes according to the focus areas of the WASH-FAST survey,

89

Results

1.WASH-FAST

In this section, we present the results of the WASH survey at the facility level

and ward level. At ward level, these results are based on the specific WASH

domains and by levels of responsibility. We then use data from the qualitative

interviews to shed light on some of the quantitative findings from the WASH

indicator survey. We excluded two hospitals from our analysis (H12 and H15).

H12 is a rural health centre and therefore not comparable to the other more

extensive facilities. In H15, which is a national referral hospital, we only

assessed the neonatal unit and thus not able to compare with the other

hospitals.

The bed capacity for the 14 participating hospitals ranged from 131 to 594 beds

and the number of wards from 5 to 14. A total of 116 (85.3%) out of 136 wards

were assessed in the survey (Table 1). Of the 20 excluded wards, six were renal;

five were psychiatric, four were Ear Nose and Throat (ENT), three were

Intensive Care (ICU) and one each was an eye and a neurosurgical unit.

90

Table 1 Summary of Hospital size and wards assessed

Facility Hospital Bed capacity

Number of specialist doctors (consultants)

Number of wards in the facility

Wards assessed

Wards evaluated by Specialty Wards Excluded

Medical Mixed Medical Surgical

Neonatal Unit

Paediatrics Surgical

High Malaria Prevalence Zone

H1 203 5 7 6 3 0 0 1 2 Renal Unit

H3 550 12 14 11 4 2 0 1 4 Psychiatry, Renal Unit, ENTa

H7 180 7 6 6 3 1 0 1 1 None H8 250 14 8 7 3 1 0 1 2 Renal Unit H14 165 5 5 5 1 3 0 1 0 None Low Malaria prevalence Zone

H2 594 26 16 12 5 1 1 1 4 ENTa,ICUb, Psychiatry, Renal

H4 216 8 7 7 3 1 0 1 2 None H5 231 7 9 8 3 1 1 1 2 Psychiatry

H6 383 17 10 9 3 2 1 1 2 Neurosurgery H9 550 19 18 14 4 2 1 2 5 ENTa,Renal,

Psychiatry, ICUb H10 131 24 6 6 2 1 1 1 1 None

H11 320 21 9 9 4 3 1 1 0 None H13 378 20 15 10 4 1 1 1 3 ENTa,Renal,

Psychiatry, ICUb, Eye H16 350 14 6 6 2 1 1 0 2 None Total 136 116 44 20 8 14 30 20

a Ear Nose and Throat b Intensive care unit

91

Performance-based on Facility-level scores

Based on the 65 indicators assessed spanning all four domains at the facility

level, we present overall performance in each hospital in Fig 2 (vertical bars).

Performance varied from 47% (H1) to 71% (H6) with a median of 61% IQR [56-

65]. The central grid in Fig 2 represents the performance of the four domains in

each hospital. Two hospitals (H4, H5) had an aggregate score of <40% in the

hygiene domain (represented by red tiles in the colour chart). The domain

scores for pooled hospital data are presented by the horizontal bars in Fig 2.

From these horizontal bars, we note the hygiene domain performed poorest at

57%, and all the overall four domain scores are below 80 %.

Fig 2. Overall WASH Performance. The overall WASH facility performance based on all 65

indicators in four domains is shown by the upper vertical bars. The right horizontal bars

summarise the performance of each domain across 14 hospitals. The tiles in the central grid are

coloured according to the performance classification of each domain in each hospital, as shown

in the colour legend.

92

Ward level Performance

The ward specific scores are represented by dots in the scatter plot in Fig 3. A

median within hospital ward score (blue vertical lines in Fig 3) based on an all

hospital’s ward specific scores was generated for each of the WASH domains.

These median domain scores were all less than 60% (water domain 59%,

IQR[48-67], sanitation 55%, IQR[46-59], hygiene 54%, IQR[41-62] and

organizational management 40%, IQR[30-60]). There was notable variability in

the ward performance within hospitals, most marked in the water domain in

hospital H9 in which 14 wards were assessed (range 40-90%). To compare this

ward level performance with the overall score in the same facility, we include

the overall facility level for each domain in each hospital (represented by non-

shaded circles in Fig 3). By arraying the individual ward performance within

each WASH domain for each hospital, we note clusters of under-performance

mainly in the sanitation and management domains, where variability is

substantial and outlier wards. The substantial differences between facility-level

performance and ward performance are attributed to the inclusion of

assessments spanning service areas in the overall facility score not captured in

ward-based scores. We provided as an appendix the individual ward level

performance (minimum and maximum score, mean, median and IQR) for each

of the 34 indicators (S1 Table).

93

Fig 3. Horizontal scatter plot of the aggregate ward level scores (black points) by domain. (O) is overall facility aggregate score for each hospital by domain.

blue vertical line (median ward score for that domain). The colour bars represent cut off values of red <40%, orange 41-60%, yellow 61-80% and green 81-

100%.

94

WASH Domain Performance

Here, we examine how the individual indicators in the domains performed at

the ward level. In Fig 4, for each specific indicator in the sanitation domain, we

present the mean performance for each hospital (colour chart in the central

grid). Also shown is the mean performance for each indicator across all the 116

wards (the horizontal bars in Fig 4) and the mean ward performance in the

sanitation domain for each hospital (vertical bars in Fig 4). The WASH indicators

that performed poorly in H11 (indicated by red tiles under H11 in the central

colour grid) are waste management standard operating procedures, waste

segregation, and availing toilet handwashing stations and cleaning records. We

provide similarly constructed figures for the water, hygiene and organisational

management domains in S4 File. Drawing from these additional figures, other

poorly performing areas were drinking water storage (8%) and ensuring a

minimum 2.5-metre distance between hospital beds in the hygiene domain

(11%).

95

Fig 4. WASH performance for sanitation domain at ward level. The vertical and horizontal bars summarise the performance of each indicator across the hospital wards. The tiles in the central grid are coloured according to the performance

96

Ward Level Performance by levels of responsibility

Ward level data are also presented based on levels of responsibility. At ward

level, we focus on two primary levels, indicators that need the action of senior

hospital managers (16 indicators) and those reflecting activities the

management can delegate to an infection prevention and control committee

(16 indicators). The overall hospital performance for the indicators under the

hospital management was between 44-65% with a median score of 56% IQR[36-

57]. Of the 16 indicators, 10 had a score below 60% with poor performance

noted on indicators that involved infrastructure (built environment). These

include the availability of hand hygiene stations at points of care(53%) and

service areas(51%), availability of connected taps(58%) and stations for drinking

water(34%)and storage(8%). For the indicators under the IPC committee, the

overall score ranged from 30-70% with a median score of 45% IQR[36-57].

There were 11 indicators with a score of less than 60%. The figures describing

these indicators are provided as a supplement( S 5 File)

2. Understanding Variation – Qualitative Analysis

1. Challenges with the built environment

Many of the facilities we surveyed are more than 40 years old and have not

been renovated or modernised, or if they have been, this has been done in a

partial way that affects only individual wards or areas of the hospital. This leads

to variation both between and within facilities. The newer hospitals and ward

units had more sinks, and toilets available, and their arrangement into cubicles

compared to an open floor plan improved bed spacing. The workers seemed to

take such arrangements as fixed and beyond their control, but health workers

recognised the importance of the built environment:

97

“The sinks like the ward I am in, and I have seen the other

ones as well, there are sinks in every cube, that is about eight

beds, there is a sink in every eight-bed a sink.” Health worker

“Right now, I talk of where I am working in the female ward

and the dressing room, we are missing the sink between the

beds. We need a sink for cleaning hands in that place. Right

now, there is no sink.” Health worker

The built environment and the quality of the facilities enable or retard progress

in WASH. For both facilities that scored red in ‘Hygiene’ (Fig 1), the facilities

were outdated and difficult to keep clean. Not having sinks available makes life

difficult for these facilities, and the results of the survey were reflected in

respondents’ comments on potential underperformance:

“But how we do it in [H5] we don’t really have, you know, a

tap everywhere and soap everywhere.” Nurse

In addition to the overall condition of some wards, the design and layout of

hospitals also seemed to play a part. Some of the buildings in these hospitals

were built more than 40 years ago at a time when essential IPC and WASH

standards were perhaps not prioritised and were constructed with small

populations in mind. Therefore, the regional governments and hospital

managers have inherited hospitals whose size and structure are indeed not

within their control. This was described by one of the regional government

leaders who highlights challenges with space for waste collection.

“Another challenge I want to say is space because you will find

that in a good set up, we need to have a space for collecting

the waste that is specifically labelled waste collection area.

98

Whereby the officers who transport the waste from the

generating area to the disposal area, should have a

designated area which is not in our hospital at the moment,

we are just improvising. But in most places, that are well

constructed, they are places you will even lock so that the

issue of contamination is minimal. In this county except for the

new lab that has a designated waste collection area.”

Regional government officer

2. Resource availability and allocation

The budgets available for IPC and WASH very much varied by region. Some of

the facilities in the survey reported challenges with budgetary allocation at the

county government level for provision of WASH materials leading to shortages

in supply or having sub-standard materials, as highlighted by one of the IPC

hospital leaders whose facility had to contend with ill-fitting personal protective

equipment.

“The challenge is inadequate funding now that you are given

some money that cannot meet your expectations then that

one comes as a challenge because perhaps you want to buy a

certain amount of material, but because of limited funding

you buy as per what is available.” Health Manager

Importantly, IPC rarely had a specific budget, and our respondents explained

how this made it challenging to ensure supplies:

“I would talk about maybe the sanitary requirement that we

need to help us do prevention but not IPC budget as per se.

There is nothing like that.” Frontline healthcare worker

99

This is important because it results in splitting the responsibilities for ensuring

regular supplies across several areas and individuals:

“You know normally we are told to itemise whatever we

require in the departments that we are working. Like now the

public health officer is the one who deals with the sanitation

and such issues. [But] chlorine it is under the pharmacy, it is

under non-pharmaceutical commodities, so we give her, the

pharmacist a figure or the requirements of the consumption in

the institution and he or she factors that.”

Frontline healthcare worker

This may also speak to some of the variations we see between wards in a single

facility where some individual ward managers are better at procuring funds for

IPC materials than others (see Fig 3).

3. Leadership at hospital and ward level

In some of the facilities, hospital managers did not appear to see WASH/IPC as

a priority area, and there was no institutional leadership or commitment to

tackle WASH challenges.

“you know change of the management brings a tug of war,

and people normally have their own ...there is what they

value. What I want basically to tell you people have not taken

infection prevention as a key concept in health sector they

take lightly ….when you come to certain managers, there are

those who don't value infection prevention they see it as by

the way.” IPC Lead

100

In facilities that focused more on WASH, attempts were sometimes made to get

better representation from the ward level leaders, such as hospital consultants

and senior nurses, to take leadership roles in IPC committees to mobilise

broader support for WASH. These doctors and senior nurses seemed vital in

improving WASH at the ward level.

“like now … we have a paediatrician who is now chair-

chairing our committee for infection prevention, but he’s

working in the paediatric unit. .. That is now what makes it

easier because now for the doctors… if there is something that

now the doctors needed to be communicated to, we use him.”

Ward Manager

Local leadership can also affect hospitals and result in variable performance

between wards, as seen in Fig 3. These wards were often those where patients

were understood to be vulnerable such as the NBU and had persons who

enforced improved IPC behaviour:

“I think it is because we have somebody who is very vocal. And

she is very strict. She will tell them you are not going to handle

any baby without having washed your hands.”

Health Manager

A manager in one of the better-performing facilities, noted the positive effects

of having such a “champion” at the facility level, leading the overall approach

and IPC committee:

“I think the nurse the in charge – the unit matron who is in

charge of that department [ IPC/ WASH] – is somebody very

101

passionate about the department, and more so you’ll find

even she keeps on pushing us.” Hospital Deputy director

3. Improving Wash – ideas from the field

In addition to the critical areas of facilities, budget and leadership, we also

identified other key insights from the qualitative interviews that affect WASH

performance at ward and facility levels. These would be key areas to consider

for intervention to improve WASH and IPC in hospitals

1. Outsourcing – a solution for general cleanliness?

With the approval of the county governments, to combat challenges of staff

shortages, supplies and poor accountability, some of the hospitals had begun

to outsource cleaning services. These companies, at a fixed price contract,

provided personnel and supplies and were responsible for ensuring these

facilities were clean at all times.

“the cleanliness in this place in comparison with most of the

other you know public institutions in this country is different

more so the people you see sweeping; cleaning are not

employees of the ministry they’re not employees of the

hospital we have outsourced it to somebody who has been

doing it in other various places.” Deputy Hospital Director

This strategy did not seem to work in all facilities as some of the poorly

performing hospitals like H11 also had outsourced cleaning services. It may still

come down to the hospital management to ensure these companies are

working well.

102

2. Improving personal and professional attitudes towards WASH

and IPC

Compared to other areas in the hospital, WASH and IPC are not taken as

seriously, mainly because these are not revenue-generating activities. These

have also been viewed historically as the premise of some particular cadres like

nurses. Therefore, doctors and other medical specialists have found it difficult

to take orders from WASH focal persons who may be junior to them.

“…sometimes when you go for this IPC meetings, and

something on IPC comes up you could feel some staff

complaining, especially the nurses that doctors don’t embrace

the issues of IPC. The other issue also is the attitude of people;

some people have very bad attitudes. They imagine I’m

being... it’s like you are bothering them.” Nurse Manager

The nurses we spoke to also noted that it was sometimes hard to persuade

other cadres of the importance of handwashing:

“I think they[Nurses]are more conversant with the infection

prevention issues, and they observe the protocols but the

other cadres, they don’t observe. I think they are not. Maybe

they are not updated as to why they should observe the

protocols and they, by the way, I would say, they don’t mind

about the way they do their things.” Nurse manager

In contrast, other hospitals, the perspective is different and more positive.

Some health workers see WASH and IPC as key to the provision of care in the

hospital.

103

“IPC is very important because I think to me it’s the heartbeat

of the hospital.” Frontline Health worker

3. Training and orientation of all cadres of staff on WASH

To correct some of the poor attitudes and challenges of WASH and IPC. Some

hospitals have taken up the training of all cadres to WASH and IPC as part of the

hospital orientation to ensure everyone is up to date.

“its routine that like now when you get posting all people

nurses and clinical officers and medical officers’ interns when

they come here now, it has become even before they get into

the ward that’s [orientation on IPC]the first place they step in

then they are taken through, and it has become very much

entrenched... in their absorption into this place.”

Deputy Medical Director

4. Partnerships to improve WASH offer partial improvement in

the sector

Some hospitals, due to constraints in funding, have linked with development

partners to provide training (e.g. H10 and H16), assist in the development of

standard operating procedures and promotional materials. Although these

partners have a positive effect, we noted in some hospitals, duplication of roles,

for example, multiple different partners engaged in promoting handwashing.

Like for the [partner X] they had the concept, they taught us

about their handwashing, the decontamination, the waste

segregation the storage and about the how to care for the

laundries and all that. All those steps were taught but [partner

104

Y] they came in the same, but they narrow to the

handwashing. Hospital manager

Discussion

Here we report the combined hospital performance of the four WASH domains

in 14 hospitals. Performance varied at hospital, ward and indicator level with

most being sub-optimal. Some of the themes explaining this variation and

performance included differences and challenges within the built environment,

resource allocation, leadership, training on WASH and health worker attitudes

towards WASH and IPC.

The JMP also assessed some of the indicators assessed in WASH-FAST in the

2016 global assessment of WASH in health facilities [4]. Although using different

methodologies, our results confirm and highlight similar issues. The main areas

showing the same poor performance with the JMP are the WASH infrastructure,

including water and toilet availability and health care waste management. The

JMP and other related WASH surveys have shown slow progress in the

improvement of WASH in most hospitals in resource-limited settings. This is

despite improving WASH in health care facilities being highlighted as one of the

sustainable development goals [13].

WASH in health facilities is a crucial intersection point for many hospital-based

interventions. These include IPC, control of antimicrobial resistance, improving

the safety and quality of care and health system strengthening interventions all

aimed towards improving efficiency and patient outcomes in hospitals. A focus

on improving WASH would, therefore, have a positive effect on multiple aspects

of health care. Some of the main inputs needed to improve WASH in health

facilities are political resources, financial/material resources and human

resources [2]. The leading players to provide these inputs are the national and

105

regional governments, policymakers and at hospital level the senior hospital

managers who may also delegate to specific committees.

Political resources include engagement from the national and county

governments and development partners. These provide national

standards/guidelines and accountability mechanisms to ensure that facilities

are meeting these national standards [14]. From our survey, the results and

findings that that would be relevant to national/regional governments are the

findings at the overall facility level. These show varied hospital performance

across the country. The role played by the government, in this case, would be

to ensure all hospital achieve a minimum standard for WASH. For example, a

directive from the national government on standards for building hospitals

would provide minimum standards for WASH infrastructure are present in

every facility especially for newer hospitals and in older hospitals that may be

undergoing renovation. These governments are also crucial in positioning

WASH on the national health agenda linked to the sustainable development

goals, universal health coverage [2] and the prevention of AMR to help make

WASH a priority for all.

Inadequate resource allocation was highlighted as a challenge affecting the

performance of WASH in these facilities. To improve WASH, there needs to be

resource allocation from the national/county government with budgets for

policy generation, assessments, upgrading of infrastructure and training for

WASH. At the hospital level, to improve and maintain the WASH infrastructure

and practices, resources are required for ongoing staff training supported by an

IPC/AMR committee, repairs need to be prioritised, and a constant supply of

quality materials should be ensured with regular assessments of all the wards

performance. Resources for WASH assessments and rewards to staff for best

performing wards may also need to be allocated [14, 15]. In some of the

106

hospitals, some challenges were overcome by partnering with development

agencies to provide extra resources to improve WASH in the hospitals. Hospital

and ward level WASH data as provided in this survey would be essential to

highlight the main areas or indicators within the hospital or wards that need to

be prioritised during resource allocation. We noted at ward level in these

hospitals that there was poor performance for some critical indicators that

would be quick and less costly to improve. These include; availing a cleaning

record in the wards (15%) and toilets (27%) and providing waste management

standard operating procedures (SOP) (30%) and hand hygiene promotion

materials(42%).

Human resource allocation within hospitals is also crucial to improving WASH.

When these staff, including cleaners and waste handling personnel, are well

trained and motivated, they will provide better service. All the workers,

whether employed or contracted need to be adequately trained and oriented

on WASH and IPC. In some of the hospitals we assessed, this process of training

and orientation was now routine. Training on WASH can be formal preservice

or in-service training and mentoring. To achieve these levels of training and

competence in these hospitals, there is a need for some trained WASH focal

persons to provide technical expertise in hospitals [2]. These focal persons

might be recognised through certification, as in the case in some African

countries [6]. In facilities with staff shortages, outsourcing of some of these

workers from cleaning companies could improve efficiency, help hospitals focus

on their core business of providing quality healthcare and may contain costs

[16]. The hospital managers are, however, ultimately responsible for the

performance of these outsourced services and therefore need to monitor the

services provided actively, tools such as WASH-FAST could help this process.

107

This survey also highlights the role of committed leadership in improving WASH.

Leadership at all levels of the health systems is an essential contributor to

improving WASH and averting AMR. National leadership provides a clear road

map for the country to follow and aligns all WASH/AMR efforts to achieve the

set targets. It should offer tools and approaches for assessment of WASH/AMR

in the country, which should be prioritised over those provided by other donor

partners to ensure coherence [14]. Leadership at the hospital level is also

critical in improving WASH/AMR. Hospital managers are vital to improving staff

morale and attitudes to embrace WASH as part of their core functions, for

example, by including these functions in the staff appraisal. Hospital

management can also delegate some of the WASH activities to committees

within the hospital. In this case, a well-constituted IPC committee would be

essential in offering training on WASH, IPC and AMR, and conducting periodic

assessments of the hospital and offering technical and budgetary

recommendations to the senior hospital managers[8]. Including senior doctors

and nurses who actively participate in these committees can enhance their and

effectiveness credibility and improve attitudes clinicians have towards

WASH[17]. Having champions of change (WASH champions) at the ward level

for activities like hand washing has previously been noted to generate positive

change [15]. This approach of using champions may also prove valuable in

improving the overall outlook of WASH in healthcare facilities, as shown in

some of the hospitals surveyed.

Financing for health in many countries, including Kenya remains a challenge.

The current budget for health below 8% of the total national budget, and this is

still below the global recommendations.[18]. This proportion has not increased

for several years despite new challenges like AMR emerging in recent years. In

Kenya with a devolved health system, the amount of money that would be set

108

aside for activities like buying IPC and WASH materials and other non-

pharmaceutical agents was only 5% of the budget with more than 70% of the

county budgets for health going to personnel costs.[18] Therefore, to improve

WASH, IPC and the AMR agenda, especially in public hospitals, there is a need

to examine other sources of funding for these activities. Public-private

partnerships, donor agencies may provide alternative ways to fund these

activities in limited-resource settings. In Kenya, the National hospital insurance

fund, which is the national health insurance provides reimbursement to public

and private facilities for some of the costs incurred by some of its members.

Pegging some of these reimbursements to the quality of care provided,

including the state of IPC and WASH in these hospitals may force these facilities

to improve[19].

From our work using the WASH-FAST tools, we demonstrate that; surveys can

be carried out at national level or regional levels to generate data on the state

of WASH in healthcare facilities in countries. These data have the potential for

use in priority setting for WASH interventions, policy generation and resource

allocation both at the national and hospital level. These data generated at the

hospital level can also be used by authorities for hospital accreditation and

benchmarking.

The main limitations of our survey were that the WASH-FAST assessment tool

provided for only three possible outcomes for each indicator. This may have

implications in the accuracy/objectivity of the score presented, to mitigate this,

we ensured all the clerks were well trained before the survey, and the data

collection team provided notes against every indicator to ensure the score

provided was as objective as possible. The other limitation in our approach was

conducting the surveys and interviews over almost the same period. Although

interviews were conducted after the survey teams shared their daily findings

109

with the social scientists and some of the social scientists were also present

during the surveys, a more sequential approach would have been preferable.

That is, after completing and analysing the quantitative data, more relevant

interview questions might have been formulated, tested, and interviews might

then have been conducted across most survey hospitals. Recruiting health

workers from participating hospitals to collect data was likely to introduce bias.

To mitigate, we ensured that these data collection teams were a mix of health

workers from the participating hospital and others from different hospitals and

included a study team leader. The indicator scores assigned were also arrived

at by team consensus. The main aim of including health workers from

participating hospitals was to build capacity for these facilities to carry out

similar surveys in future as part of their facility improvement strategies.

Future work using the WASH-FAST might explore how we could systematise

feedback to hospitals of survey results and recommended follow-up actions,

including the use of electronic dashboards. Currently, all the indicators have

equal weight, but we acknowledge some may be more critical for patient safety

or in the prevention of acquired infections. Future work would, therefore,

consider ways to give more weight to more critical indicators during surveys,

analysis, reporting and follow-up.

Conclusion

Significant differences and challenges exist in the state of WASH within and

across even large hospitals providing multi-speciality care. Whereas the senior

hospital management can make some improvements, input and support from

the national and county governments are essential to improve WASH as a

necessary foundation for safe, quality hospital care and to avert AMR in Kenya.

110

Acknowledgements

The authors thank the Kenyan Ministry of Health and the Council of Governors

who permitted this work to be carried out. The authors thank Dr Nancy Abuya

and all the clinicians who assisted in data collection. We also thank the hospital

management and clinical teams who supported the work in the survey

hospitals. The authors also thank Arabella Hayter and Margaret Montgomery

from the Water Sanitation, Hygiene and Health Unit, Department of Public

Health and the Environment, WHO for providing WASH training materials and

technical advice. This work is published with the permission of the Director of

KEMRI.

Author contributions

The roles of the contributors were as follows: MM, J.M, O.T, C.S and M.E

conceived the study. M.M, G.K and M.Z, collected data, P.M, O, T, J.M assisted

M.M in data analysis and interpretation. M.M drafted the manuscript. M.M,

J.M, O.T, C.S and M.E critically revised the manuscript for intellectual content.

All authors read and approved the final manuscript.

Data Availability

All summary data underlying the findings are freely available in the

manuscript and supplemental files. The raw data used for this manuscript are

hosted in a public repository Harvard Data verse. DOI

https://doi.org/10.7910/DVN/IJUWWR

111

References 1. Hans Jørn Kolmos. Health Care Associated Infections: Sources and Routes of

Transmission. In: Infection Control – Updates Croatia2012 [cited 2017 2nd November ]. Available

from: http://cdn.intechopen.com/pdfs-wm/28876.pdf.

2. Weber N, Patrick M, Hayter A, Martinsen AL, Gelting R. A conceptual evaluation

framework for the water and sanitation for health facility improvement tool (WASH FIT). Journal

of Water, Sanitation and Hygiene for Development. 2019. doi: 10.2166/washdev.2019.090.

3. World Health Organisation. Safe management of waste from health care activities

2014 [cited 2018 18th December ]. Available from:

http://www.who.int/water_sanitation_health/publications/wastemanag/en/.

4. World Health Organization & United Nations Childrens Fund (UNICEF) WASH in health

care facilities: global baseline report 2019. Geneva: 2019.

5. World Health Organization, UNICEF. Water and Sanitation for Health Facility

Improvement Tool (WASH FIT): a practical guide for improving quality of care through water,

sanitation and hygiene in health care facilities. Geneva: 2017 9241511699.

6. Adams J, Bartram J, Chartier Y. Essential environmental health standards in health care:

World Health Organization; 2008.

7. World Health Organization. Guidelines on core components of infection prevention and

control programmes at the national and acute health care facility level: World Health

Organization; 2016.

8. Maina M, McKnight J, Tosas-Auguet O, Zosi M, Kimemia G, Mwaniki P, et al. Extending

the use of the World Health Organisations’ Water Sanitation and Hygiene assessment tool for

surveys in hospitals – from WASH-Fit to WASH-FAST. 2019:Forthcoming.

9. English M, Ayieko P, Nyamai R, Were F, Githanga D, Irimu G. What do we think we are

doing? How might a clinical information network be promoting implementation of recommended

paediatric care practices in Kenyan hospitals? Health Research Policy and Systems. 2017;15(1):4.

doi: 10.1186/s12961-017-0172-1.

10. Weber Nicole, Martinsen Andrea L, Sani Amidou, Assigbley Elom Kokou Eric, Azzouz

Chedly, Hayter Arabella, et al. Strengthening Healthcare Facilities Through Water, Sanitation, and

Hygiene (WASH) Improvements: A Pilot Evaluation of “WASH FIT” in Togo. Health Security.

2018;16(S1):S-54-S-65. doi: 10.1089/hs.2018.0042.

11. R Core Team. R: A Language and Environment for Statistical Computing. Vienna,

Austria: 2017.

12. QSR International Pty Ltd. NVivo qualitative data analysis Software;. 12 ed2018.

13. United Nations General Assembly. Transforming our world: the 2030 Agenda for

Sustainable Development. New York: United Nations. 2015;(1).

14. World Health Organization. Water, sanitation and hygiene in health care facilities:

practical steps to achieve universal access to quality care. Geneva: 2019.

15. World Health Organisation. Global strategy, burden of disease, and evidence and action

priorities, Workshop Report. United Kingdom: 2016.

16. Ichoho JW. Implementation of outsourcing strategy at the Nairobi Hospital, Kenya.

[Unpublished MBA Project]. In press 2013.

112

17. Amirabadi Zadeh H, Maleki MR, Salehi M, Watankhah S. An Exploration of the Role of

Hospital Committees to Enhance Productivity. Global journal of health science. 2016;8(3):199-

209. Epub 2015/10/27. doi: 10.5539/gjhs.v8n3p199. PubMed PMID: 26493440; PubMed Central

PMCID: PMCPMC4804034.

18. Ministry of Health Kenya. National and county health budget analysis FY 2016/17. 2017.

19. Barasa E, Rogo K, Mwaura N, Chuma J. Kenya National Hospital Insurance Fund

Reforms: Implications and Lessons for Universal Health Coverage. Health Systems & Reform.

2018;4(4):346-61. doi: 10.1080/23288604.2018.1513267.

Supporting Information

S1 Table. WASH Aggregate Ward Indicator Scores

S2 File. WASH Data collection and Standard Operating procedures tool

S3 File. Consolidated criteria for reporting qualitative studies (COREQ): a 32-

item checklist

S4 File Ward level Aggregate by Domain for Water, Hygiene and Organisation

Management

S5 File Performance of Indicators under the infection prevention and control

committee and hospital management

113

S1 Table Aggregate Ward Indicator Scores

DOMAIN Indicator Description Minimum

Score

lower

IQR

Median

Score

Upper

IQR

Maximum

Score

Mean

Score

Water

Showers available 58.33 75.00 84.82 92.86 100 83.62

Shower lighting present 16.67 44.18 60.63 69.17 96.43 56.99

Water Services Available in quantity 63.64 92.71 100 100 100 93.12

Drinking water Station 0 0 17.14 72.57 100 34.05

Drinking Water Storage 0 0 3.13 13.49 28.57 7.96

End points(taps) connected 0 46.88 54.91 78.75 91.67 58.19

Sanitation

Waste collection bins 50 50 57.74 66.25 100 62.92

Waste segregation 0 12.50 50 72.92 94.44 45.32

Waste management SOP 3.57 13.54 21.88 42.56 80 29.98

Waste handling PPE 28.57 50 60.42 71.67 100 60.99

Toilets available in correct numbers 55.00 66.67 71.43 88.33 100 76.23

Toilets well lit 16.67 37.20 58.12 65.00 91.67 54.48

Toilets separate (Male/ Female,

Staff/Patients)

42.86 74.55 92.92 100 100 84.73

Toilet with Menstrual Hygiene services 20 64.88 83.33 89.38 100 73.28

Toilets for reduced Mobility 0 0 0 4.69 33.33 4.91

114

Toilet Hand wash station 28.57 42.86 50 53.13 75.00 49.20

Toilet cleaning record 0 0 0 50 95.00 26.64

Hygiene

Insecticide treated nets on beds 0 0 24.70 54.69 83.33 29.51

Mechanism for IPC supplies tracking 0 17.86 81.94 100 100 63.41

Cleaning records available 0 0 0 0 90 14.52

Natural ventilation available 65.00 94.27 100 100 100 95.14

2.5 Metre bed distance 0 0 4.17 18.85 41.67 11.31

General lighting adequate 58.33 95.55 100 100 100 95.44

Clean floors/surfaces 50 70.31 87.50 100 100 83.79

cleaning materials available 41.67 50 56.70 64.58 83.33 58.12

Hand Hygiene-Points of care 35.71 50 51.79 56.08 70 53.35

Hand Hygiene Promotion materials 0 33.33 43.30 50 90 41.67

Hand Hygiene-Service areas 18.75 50 50 50 77.78 50.88

Hand Hygiene compliance activities 0 0 45.00 53.41 80 34.99

Organisation

Management

Staff rewards for good performance 0 0 26.70 50 61.11 25.78

Ward based audits 0 21.97 36.61 48.96 70 35.28

New staff orientation 43.75 78.33 92.26 100 100 87.06

Staff IPC training 0 19.32 35.83 50 100 37.89

IPC Focal person 0 2.08 20.83 48.44 91.67 31.88

115

S2 File. WASH Data collection and Standard Operating procedures tool

WASH ASSESSMENT BOOKLET

WASH TEAM LIST

List all persons that have contributed to the assessment or have participated

in briefings/discussions during the WASH baseline assessment, each day

Date Name Job Title

/ Role

Role / Involvement During

WASH Baseline Assessment

Contact Details

(e.g. phone

Number and e-

mail if

available)

116

Additional Comments on Roles and Contributions during WASH Baseline Assessment

117

Notes from Initial Briefing

During the initial briefing, you must hand in appropriate approval letters, explain the purpose of the survey in the context of

the broader study, explain what the assessment will involve, and when and how feed-back will be provided. You must explain

that the assessment will involve taking pictures, to assist with the scoring of indicators in consultation with the research teams

operating in other hospitals. Pictures will NOT include patients or staff, or signs that can identify the hospital. They will be kept

in anonymised format and encrypted and will be destroyed as soon as the indicators are unequivocally scored. ONLY THE THREE

WASH TEAM SUPERVISORS (i.e. NOT the team leaders) ARE ALLOWED TO TAKE PICTURES. During the briefing, take the

opportunity to ask questions to help plan your visit: ask about numbers of wards, names of the wards, structure of the hospital,

and permanent and/or temporary local WASH and/or IPC improvement activities by NGOs or other, which may affect the

interpretation of your findings. Ask questions around hospital demographics and capacity, such as total number of beds (i.e.

hospital capacity), numbers of outpatients on a typical day, average number of surgical procedures daily or weekly, total number

of admissions daily or weekly, as well of average number of births in a day or week. You must write down the notes from the

meeting in the sections below. If you gather relevant details in subsequent days, which refer to hospital structure, capacity and

demographics, add the information here as well.

118

WARD-LEVEL

ASSESSMENT FORMS

Explanatory Notes for Ward-Level WASH Indicators

WATER

1.2 For an intermittent piped-water supply, e.g. available 8 hours per day.

1.3 Describe if drinking stations are provided with railings and a seat.

1.6 Describe how many sinks have functioning pipes for water coming in and going out and how many sinks have functioning pipes for

water coming in only.

SANITATION

2.1 Improved sanitation facilities include flush toilets into managed sewer or septic tank and soakaway pit, VIP latrines, pit latrines

with slab, and composting toilets.

To be considered usable, a toilet/latrine should have a door which is unlocked when not in use (or for which a key is available at

any time) and can be locked from the inside during use, there should be no major holes in the structure, the hole or pit should not

be blocked, water should be available for flush/pour flush toilets, and there should be no cracks, or leaks in the toilet structure. It

should be within the grounds of the facility and it should be clean as noted by absence of waste, visible dirt and excreta and

insects.

2.4 Toilets should have a bin for disposal of waste or an area for washing, with water available.

119

2.5 A toilet can be considered to meet the needs of people with reduced mobility if it meets the following conditions: can be accessed

without stairs or steps, handrails for support are attached either to the floor or sidewalls, the door is at least 80 cm wide, the

toilet has a raised seat (between 40-48 cm from the floor), a backrest and the cubicle has space for circulation/manoeuvring

(150x150 cm). The sink tap, and water outside should also be accessible and the top of the sink 75 cm from the floor (with knee

clearance). Switches for lights, where relevant, should also be at an accessible height (max 120 cm). All specifications are based on

ISO 21542:2011 – Building construction - Accessibility and usability of the built environment, available at:

http://www.iso.org/iso/home/store/catalogue_tc/catalogue_detail.htm?csnumber=50498

2.6 A functional hand hygiene station may consist of soap and water with a basin/pan for washing hands. Water should not be

chlorinated. Alcohol-based hand rub is not suitable for use at latrines.

2.7 For low literacy or illiterate cleaners, this should be adapted and simplified accordingly with recognizable pictures and illustrations.

2.10 Lighting for latrines is necessary in all facilities where night-time services are provided and where there is not sufficient natural

light to safely use the latrine during the day.

HEALTHCARE WASTE

2.12 Functional means containers should not be ¾ full, be leak-proof with a lid and clearly labelled (i.e. easily distinguishable according

to a colour, label or symbol).

2.22 Protective equipment for people handling waste management includes gloves, apron and tough rubber boots.

HAND HYGIENE

3.1 Point of care is where three elements come together: the patient, the health-care workers and care or treatment involving contact

with the patient or his/her surroundings. This may include consultation rooms, operating rooms, delivery rooms AND laboratory.

Hand hygiene stations should have a sink or bucket with tap and water with soap OR alcohol-based hand-rub. There should be at

least two hand hygiene stations in a ward with more than 20 beds. Verify that water is available from the tap.

3.2 Key places include at points of care and within 5 m of latrines. You may want to write down where stations are located (and were

they are missing) in your comments box.

120

3.3 Sink or bucket with tap and water with soap OR alcohol-based hand-rub. Service areas include sterilization room (where

applicable), laboratory (where applicable) and showers. (Toilets are included under 2.7)

3.4 Tap and water with soap.

ENVIRONMENT, CLEANLINESS AND DISINFECTION

3.8 Clean as noted by absence of waste, visible dirt and excreta and insects. Environmental surfaces or objects contaminated with

blood, other body fluids, secretions or excretions are cleaned and disinfected as soon as possible using standard hospital

detergents/disinfectants.

MANAGEMENT

N/A N/A

121

Ward Assessment - Notes

Before you begin your assessment and scoring, take the time to walk through the ward and observe all the details. Speak to

staff, take pictures where needed (no patients/staff/hospital identifying signs allowed in pictures. ONLY THE THREE WASH TEAM

SUPERVISORS (i.e. NOT the team leaders) ARE ALLOWED TO TAKE PICTURES), count numbers of functional toilets, beds, sinks,

bulbs, hand washing stations, soap dispensers, etc. Inspect bins, cleaning materials, etc. Write notes in the space below. Sit

with the team and try to score some of the indicators. If needed walk through the ward again to gather mode details which will

allow you to fine-tune the scoring. In the process, write down all your questions. You can then address all your queries with a

member of staff in one go. This should allow you to complete the assessment.

ENTER WARD NAME: _____________________________

122

Ward Name:

________________________

Facility

Code:________________________

BASELINE ASSESSMENT

(Clearly tick the box that best describes your findings) Assessment Date:

1 WATER

Meets

Target

(+++)

Partially Meets

Target

(++)

Does not

Meet Target

(+)

Explain the reasoning for your scoring and

comment as appropriate. For indicators with

an asterisk, read the notes before scoring.

1.2* Water services available at

all times and of sufficient

quantity for all uses (refers

to a typical average week).

□ Yes, every day

and of sufficient

quantity

□ More than 5

days per week or

every day but

not sufficient

quantity

□ Fewer than 5 days

per week

1.3* A reliable drinking water

station is present and

accessible for staff, patients

and carers at all times.

□ Yes, at all times

and accessible to

all

□ Sometimes, or

not available for

all users

□ Not available

1.4 Drinking water is safely

stored in a clean bucket/

tank with cover and tap.

□ Yes □ Only some

drinking water

points are safely

stored

□ Not safely stored in

any water points or

no drinking water

available

1.6* All end points (i.e. taps) are

connected to an available

and functioning water

supply.

□ Yes, all are

connected and

functioning

□ More than half

of all endpoints

are connected

and functioning

□ No, less than half

of all endpoints

connected and

functioning

123

1.13 At least one shower or

bathing area is available per

40 patients in inpatient

settings and is functioning

and accessible.

□ Yes □ Showers

available, but no

water or in

disrepair or

showers

available but

fewer than 1 per

40

□ No showers.

1.14 Shower(s) are adequately

lit, including at night.

□ Yes □ Lighting

infrastructure

exists, but not

functioning

□ Not adequately lit

or no lighting

infrastructure

Ward Name:

________________________

Facility

Code:________________________

BASELINE ASSESSMENT

(Clearly tick the box that best describes your

findings)

Assessment Date:

2.a SANITATION Meets

Target

(+++)

Partially Meets

Target

(++)

Does not

Meet Target

(+)

Explain the reasoning for your scoring and comment

as appropriate. For indicators with an asterisk, read

the notes before scoring.

2.1* Number of available and

usable toilets or improved

latrines for patients.

□ One for

every 20

inpatients.

□ Sufficient

number present

but not all

functioning or

insufficient

number.

□ Less than 50% of

required number

of latrines

available and

functioning.

2.2 Toilets or improved latrines in

the ward clearly separated for

staff and patients

□ Yes.

N/A

□ No separate

latrines.

124

2.4* At least one toilet or improved

latrine provides the means to

manage menstrual hygiene

needs.

□ Yes. □ Yes, but toilet

is not clean or

in disrepair.

□ No. (Score this question in female wards only)

2.5* At least one toilet in the ward

meets the needs of people

with reduced mobility.

□ Yes. □ Yes, but not

available or in

disrepair.

□ No toilet for

disabled users.

2.6* Functioning hand hygiene

stations within 5 m of latrines.

□ Yes. □ Present, not

functioning or

no water or

soap.

□ Not present.

2.7* Record of cleaning visible and

signed by the cleaners each

day.

□ Yes. □ Toilets

cleaned but not

recorded.

□ No record /

toilets cleaned less

than once a day.

2.10* Latrines are adequately lit,

including at night.

□ Yes □ Lighting

infrastructure

exists, but not

functioning.

□ Not adequately

lit or no lighting

infrastructure.

Ward Name:

________________________

Facility

Code:________________________

BASELINE ASSESSMENT

(Clearly tick the box that best describes your

findings)

Assessment Date:

2.b HEALTHCARE WASTE Meets

Target

(+++)

Partially Meets

Target

(++)

Does not

Meet Target

(+)

Explain the reasoning for your scoring and comment

as appropriate. For indicators with an asterisk, read

the notes before scoring.

125

2.12

*

Functional waste collection

containers for 1) non-infectious

(general) waste, 2) infectious

waste and 3) sharps waste in

close proximity to all waste

generation points.

□ Yes. □ Separate bins

present but lids

missing or more

than ¾ full; only

two bins (instead

of three); or at

some but not all

waste generation

points.

□ No bins or

separate sharps

disposal.

2.13 Waste correctly segregated at

all waste generation points.

□ Yes. □ Some sorting

but not all

correctly or not

practiced

throughout the

ward.

□ No sorting.

2.21 Protocol or SOP (Standard

Operating Procedure) for safe

management of healthcare

waste clearly visible and legible.

□ Yes,

visible and

implemen

ted.

□ Written but not

visible or

implemented.

□ No protocol/SOP

in place.

2.22

*

Appropriate protective

equipment for all staff in charge

of waste disposal.

□ Yes. □ Some

equipment

available, but not

for all staff, or

available but

damaged.

□ None available.

126

Ward Name:

________________________

Facility

Code:________________________

BASELINE ASSESSMENT

(Clearly tick the box that best describes your findings) Assessment Date:

3.a HAND HYGIENE

Meets

Target

(+++)

Partially Meets

Target

(++)

Does not

Meet Target

(+)

Explain the reasoning for your scoring and

comment as appropriate. For indicators with

an asterisk, read the notes before scoring.

3.1* Functioning hand hygiene

stations are available at all

points of care.

□ Yes. □ Stations

present, but no

water and/or

soap or alcohol

hand-rub

solution.

□ Not present.

3.2* Hand hygiene promotion

materials clearly visible and

understandable at key

places.

□ Yes. □ Some places

but not all.

□ None.

3.3* Functioning hand hygiene

stations are available in-

service areas.

□ Yes. □ Stations

present, but no

water and/or

soap or alcohol

hand-rub

solution.

□ Not present.

3.5 Hand hygiene compliance

activities are undertaken

regularly.

□ Yes. □ Compliance

activities in

policy, but not

carried out with

any regularity.

□ No compliance

activities.

127

Ward Name:

________________________

Facility

Code:________________________

BASELINE ASSESSMENT

(Clearly tick the box that best describes your findings) Assessment Date:

3.b ENVIRONMENT,

CLEANLINESS AND

DISINFECTION

Meets

Target

(+++)

Partially Meets

Target

(++)

Does not

Meet Target

(+)

Explain the reasoning for your scoring and

comment as appropriate. For indicators with

an asterisk, read the notes before scoring.

3.7 General lighting sufficiently

powered and adequate to

ensure safe provision of

healthcare including at night.

□ Yes, always. □ Yes,

sometimes.

□ Never.

3.8* Floors and horizontal work

surfaces appear clean.

□ Yes. □ Some floors

and work

surfaces appear

clean, but others

do not.

□ Most and/or all

floors and surfaces

are visibly dirty.

3.9 Appropriate and well-

maintained materials for

cleaning (i.e. detergent, mops,

buckets, etc.) are available.

□ Yes. □ Yes, available

but not well

maintained.

□ No materials

available.

3.12 Beds have insecticide treated

nets to protect patients from

mosquito-borne diseases.

□ Yes, on all

beds.

□ Available on

some but not all

beds, or available

but with rips and

or holes.

□ No bed nets

available.

128

3.13 A mechanism exists within the

ward to track supply of IPC-

related materials (such as

gloves and protective

equipment) to identify stock-

outs.

□ Yes. □ Mechanism

exists but is not

enforced.

□ No mechanism

exists.

3.14 Record of cleaning visible and

signed by the cleaners each

day.

□ Yes. □ Record exists

but is not

completed daily

or is outdated.

□ No record of floors

and surfaces being

cleaned.

3.16 The facility has sufficient

natural ventilation and where

the climate allows, large

opening windows, skylights

and other vents to optimize

natural ventilation.

□ Yes. □ Some

ventilation but

not well

maintained or

insufficient to

produce natural

ventilation.

□ No.

3.18 □ Beds for patients should be

separated by a distance of 2.5

metres from the centre of one

bed to the other and each bed

has only one patient.

□ Yes, all beds

meet this

guidance.

□ Some but not

all beds fit this

criteria.

□ No beds meet this

criteria.

Ward Name:

________________________

Facility

Code:________________________

BASELINE ASSESSMENT

(Clearly tick the box that best describes your findings) Assessment Date:

129

4 MANAGEMENT

Meets

Target

(+++)

Partially Meets

Target

(++)

Does not

Meet Target

(+)

Explain the reasoning for your scoring and

comment as appropriate. For indicators with

an asterisk, read the notes before scoring.

4.6 Regular ward-based audits by

external facility personnel (not

ward staff) are undertaken to

assess the availability of hand-

rub, soap, single use towels and

other hand hygiene resources.

□ Yes. □ Undertaken

less than once a

week or

assessment is

incomplete.

□ Not undertaken.

4.7 New healthcare personnel

receive IPC training as part of

their orientation program.

□ Yes. □ Some but not

all staff.

□ No training.

4.8 Healthcare staff are trained on

WASH/IPC each year.

□ Yes. □ Staff are

trained but not

every year or

only some staff

are trained.

□ No training.

4.9 The ward has a WASH or IPC

focal person.

□ Yes. □ Yes, but focal

point does not

have sufficient

time, resources

or motivation to

carry out duties.

□ No.

4.11 High performing staff in the

ward are recognized and

rewarded and those that do not

perform are dealt with

accordingly.

□ Yes. □ Either high or

low performers

addressed but

not both.

□ No action or

recognition of staff

based on

performance.

130

FACILITY

ASSESSMENT FORMS

Explanatory Notes for Facility-Level WASH Indicators

WATER

1.1 Improved water sources in healthcare settings include piped water, boreholes/tube wells, protected wells, protected springs,

rainwater, and packaged or delivered water.

This refers to the water supply for general purposes, including drinking, washing and cleaning.

1.2 For an intermittent piped-water supply, e.g. available 8 hours per day.

1.3 Describe if drinking water stations are provided with railings and a seat.

1.8 Water needs will vary depending on the type of facility and number of patients. To calculate the facility’s water requirements,

add up the following requirements (source: WHO 2008 Essential environmental standards in healthcare) or applicable national

standards.

Outpatients (5 litres/consultation) + Inpatients (40–60 litres/patient/day) + Operating theatre or maternity unit (100

litres/intervention) + Dry or supplementary feeding centre (0.5–5 litres/consultation depending on waiting time) + Cholera

treatment centre (60 litres/patient/day).

Acceptable storage methods include clean, covered and well-maintained containers which prevent contamination from

entering and are free from any cracks, leaks, etc. Such containers should also allow for water to be extracted without hands or

other potentially contaminated surfaces from touching the water (i.e. through use of a tap).

131

1.9 Such technologies should meet one of WHO’s Household Water Treatment and Safe Storage (HWTS) performance categories

and generally involve filters, boiling, solar, chlorine (for non-turbid water) or coagulation/flocculation. Higher performing

technologies (i.e. two- or three-stars including membrane filters, UV and coagulants/flocculants) are recommended for

vulnerable groups (i.e. those with HIV or young infants) and where the specific pathogen of concern is not known. A list can be

found here: http://www.who.int/household_water/scheme/products/en/ and further information found at the WHO

Household Water Treatment site: http://www.who.int/household_water/scheme/en/

Drinking water meets WHO Guidelines for drinking-water quality (2011) or national standards:

http://www.who.int/water_sanitation_health/publications/dwq-guidelines-4/en/

1.10 Evidence of documented chlorine residuals should be available from previous testing.

1.11 Drinking water meets WHO Guidelines for drinking-water quality (2011) or national standards:

http://www.who.int/water_sanitation_health/publications/dwq-guidelines-4/en/

SANITATION

2.1 At least four toilets per outpatient setting (one for staff, and for patients: one for females, one for males, one for disabled

users). More latrines may be needed depending on the size of the facility. Improved sanitation facilities include flush toilets

into managed sewer or septic tank and soakaway pit, VIP latrines, pit latrines with slab, and composting toilets. To be

considered usable, a toilet/latrine should have a door which is unlocked when not in use (or for which a key is available at any

time) and can be locked from the inside during use, there should be no major holes in the structure, the hole or pit should not

be blocked, water should be available for flush/pour flush toilets, and there should be no cracks, or leaks in the toilet

structure. It should be within the grounds of the facility and it should be clean as noted by absence of waste, visible dirt and

excreta and insects.

2.4 Toilets should have a bin for disposal of waste or an area for washing, with water available.

132

2.5 A toilet can be considered to meet the needs of people with reduced mobility if it meets the following conditions: can be

accessed without stairs or steps, handrails for support are attached either to the floor or sidewalls, the door is at least 80 cm

wide, the toilet has a raised seat (between 40-48cm from the floor), a backrest and the cubicle has space for

circulation/manoeuvrings (150x150 cm). The sink tap, and water outside should also be accessible and the top of the sink

75cm from the floor (with knee clearance). Switches for lights, where relevant, should also be at an accessible height (max 120

cm). All specifications are based on ISO 21542:2011 – Building construction - Accessibility and usability of the built

environment, available at: http://www.iso.org/iso/home/store/catalogue_tc/catalogue_detail.htm?csnumber=50498

2.6 A functional hand hygiene station may consist of soap and water with a basin/pan for washing hands. Water should not be

chlorinated. Alcohol-based hand rub is not suitable for use at latrines.

2.7 For low literacy or illiterate cleaners, this should be adapted and simplified accordingly with recognizable pictures and

illustrations.

2.8, 2.9 No leakage from pipes nor soakaway pit, and soakaway more than 30 m from water source, with grease trap and no visible

pool of stagnant water.

2.10 Lighting for latrines is necessary in all facilities where night-time services are provided and where there is not sufficient natural

light to safely use the latrine during the day.

HEALTHCARE WASTE

2.12 Functional means containers should not be ¾ full, be leak-proof with a lid and clearly labelled (i.e. easily distinguishable

according to a colour, label or symbol).

133

2.15 Incinerator (if designed for infectious waste and not just general waste) must follow specific design requirements (e.g. use of

fire bricks/refractory bricks and mortar (vs. common building bricks) that can withstand the temperatures needed for these

incinerators (greater than 800° C). For complete burning, a dual chamber incinerator is needed that reaches temperatures

above 800° C and 1100° C, respectively. In case dual incinerators are not available and there is an immediate need for public

health protection, small scale incinerators might be used. This involves a compromise between the environmental impacts

from controlled combustion with an overriding need to protect public health if the only alternative is indiscriminate dumping.

These circumstances exist in many developing situations and small-scale incineration can be a realistic response to an

immediate requirement. For guidelines, see WHO (2014) Safe management of waste from health-care activities.

Waste may be treated off site. If so, there should be a means to confirm it is treated safely once removed from the facility

premises.

2.18 Unless a refrigerated storage room is available, storage times for infectious waste (e.g. the time between generation and

treatment) should not exceed the following periods:

Temperate climate: 72 hours in winter / 48 hours in summer.

Warm climate: 48 hours during the cool season / 24 hours during the hot season.

Fenced area protected from flooding + lined and covered pit > 30 m from water source + no unprotected healthcare waste is

observed. If waste removed offsite, both the site and the holding area (minus the pit) should comply with the above

requirements.

2.19 Placenta pits: lined or unlined depending on the geology, with slab, with ventilation pipe.

2.20 Ash pits: lined or unlined depending on the geology but must prevent leaching to the environment, with slab, bottom of pit at

least 1,5 m away from groundwater table. If water gets into the ash pit, it can leach pollutants into the soil.

2.22 Protective equipment for people handling waste management includes gloves, apron and tough rubber boots.

HAND HYGIENE

134

3.1 Point of care is where three elements come together: the patient, the health-care workers and care or treatment involving

contact with the patient or his/her surroundings. This may include consultation rooms, operating rooms, delivery rooms AND

laboratory. Hand hygiene stations should have a sink or bucket with tap and water with soap OR alcohol-based hand-rub.

There should be at least two hand hygiene stations in a ward with more than 20 beds.

Verify that water is available from the tap.

3.2 Key places include at points of care, the waiting room, at the facility’s entrance and within 5 m of latrines.

3.3 Sink or bucket with tap and water with soap OR alcohol-based hand-rub. Service areas include sterilization, laboratory,

kitchen, laundry, showers, waste zone and mortuary. (Toilets are included under 2.7)

3.4 Tap and water with soap.

NOTES – ENVIRONMENT, CLEANLINESS AND DISINFECTION (FACILITY ASSESSMENT)

3.8 Clean as noted by absence of waste, visible dirt and excreta and insects. Environmental surfaces or objects contaminated with

blood, other body fluids, secretions or excretions are cleaned and disinfected as soon as possible using standard hospital

detergents/disinfectants.

3.10 Waste disposal staff who operate the incinerator should have an apron, gloves, goggles, face mask and boots.

NOTES – MANAGEMENT (FACILITY ASSESSMENT)

4.2 The budget refers to that used for capital and operational costs. It could be from the community-management group and/or

the government, according to the policies and practices in the country.

135

Facility Assessment - Notes

Finish assessing all the wards before you go on to assess the facility. Before you begin you assessment and scoring, take the

time to walk through all areas in the facility, including the outside area, the kitchen, the laundry… inspect the incinerator

facilities, water supply and management of waste, including grey water… and check availability and count numbers of functional

latrines, toilets, lightening, drinking water stations, hand washing stations, soap dispensers, bins, cleaning material, availability

of charts and signs in all the areas of the facility other than the wards. Speak to staff, take pictures where needed (no

patients/staff/hospital identifying signs allowed in pictures; ONLY THE THREE WASH TEAM SUPERVISORS (i.e. NOT the team

leaders) ARE ALLOWED TO TAKE PICTURES). Sit with the team and try to score some of the indicators. If needed walk through

the facility again to gather mode details which will allow you to fine-tune the scoring. In the process, write down all your

questions. You can then address all your queries with a member of staff in one go. This should allow you to complete the

assessment. The assessment refers to the OVERALL scoring for the facility after having observed all wards and all areas.

ENTER FACILITY CODE: _____________________________

136

OVERALL FACILITY ASSESSMENT

ENTER FACILITY

CODE:_________________

BASELINE ASSESSMENT

(Clearly tick the box that best describes your findings) Assessment Date:

1 WATER

Meets

Target

(+++)

Partially Meets

Target

(++)

Does not

Meet Target

(+)

Explain the reasoning for your scoring and

comment as appropriate. For indicators with

an asterisk, read the notes before scoring.

1.1* Improved water supply

piped into the facility or on

premises and available.

□ Yes, improved

water supply

within facility and

available

□ Improved

water supply on

premises,

(outside of

facility building)

and available

□ No improved water

source within facility

grounds, or improved

supply in place but

not available.

1.2* Water services available at

all times and of sufficient

quantity for all uses (refers

to a typical average week)

□ Yes, every day

and of sufficient

quantity

□ More than 5

days per week or

every day but not

sufficient

quantity

□ Fewer than 5 days

per week

1.3* A reliable drinking water

station is present and

accessible for staff, patients

and carers at all times and

in all locations/wards.

□ Yes, at all

times/wards and

accessible to all

□ Sometimes, or

only in some

places or not

available for all

users

□ Not available

1.4 Drinking water is safely

stored in a clean bucket/

tank with cover and tap.

□ Yes □ Only some

available drinking

water points are

safely stored

□ Not safely stored in

any water points or

no drinking water

available

137

1.5 Sanitary inspection risk

score (using Sanitary

Inspection Form).

□ Low risk □ Medium risk □ High or very high

risk

1.6 All end points (i.e. taps) are

connected to an available

and functioning water

supply.

□ Yes, all are

connected and

functioning

□ More than half

of all endpoints

are connected

and functioning

□ No, less than half of

all endpoints

connected and

functioning

1.7 Water services available

throughout the year (i.e.

not affected by seasonality,

climate change-related

extreme events or other

constraints). Refers to ‘in a

typical average year’.

□ Yes, throughout

the year

□ Water

shortages for 1-2

months

□ Water shortages

for 3 months or more

1.8* Water storage is sufficient

to meet the needs of the

facility for 2 days.

□ Yes □ More than 75%

of needs met

□ Less than 75% of

needs met

1.9* Water is treated and

collected for drinking with a

proven technology that

meets WHO performance

standards.

□ Yes □ Treated but

not regularly

□ Not treated

138

1.10* Drinking water has

appropriate chlorine

residual (0.2mg/l or 0.5mg/l

in emergencies) or 0

E.coli/100 ml and is not

turbid.

□ Yes □ Chlorine

residual exists

but is <0.2mg/l

□ Not treated / do

not know residual

/do not have capacity

to test residual/ no

drinking water

available

1.11* National water quality

standards exist, and the

facility water supply is

regulated according to

these standards.

□ Yes, and water

meets national

standards.

□ Yes, but no

regulation or

testing.

□ No standards exist.

1.12 Sufficient energy is available

for pumping and boiling

water (mark if not

applicable).

□ Yes, always □ Yes, sometimes □ Never

1.13 At least one shower or

bathing area is available per

40 patients in inpatient

settings and is functioning

and accessible.

□ Yes □ Showers

available, but no

water or in

disrepair or

showers

available but

fewer than1 per

40

□ No showers.

1.14 Shower(s) are adequately

lit, including at night.

□ Yes □ Lighting

infrastructure

exists, but not

functioning

□ Not adequately lit

or no lighting

infrastructure

139

OVERALL FACILITY ASSESSMENT

ENTER FACILITY

CODE:_________________

BASELINE ASSESSMENT

(Clearly tick the box that best describes your findings) Assessment Date:

2.a SANITATION Meets

Target

(+++)

Partially

Meets Target

(++)

Does not

Meet Target

(+)

Explain the reasoning for your scoring and

comment as appropriate. For indicators with an

asterisk, read the notes before scoring.

2.1* Number of available and

usable toilets or improved

latrines for patients.

□ 4 or more

(outpatients)

and one per

20 users

(inpatients).

□ Sufficient

number

present but

not all

functioning or

insufficient

number.

□ Less than 50% of

required number of

latrines available and

functioning.

2.2 Toilets or improved latrines

clearly separated for staff and

patients and visitors.

□ Yes. □ Some

separate

latrines but

not for all

three

categories

(staff, patients

and visitors).

□ No separate latrines.

2.3 Toilets or improved latrines

clearly separated for male and

female.

□ Yes.

N/A.

□ No indication of

gender separation.

2.4* At least one toilet or improved

latrine provides the means to

manage menstrual hygiene

needs.

□ Yes. □ Yes, but

toilet is not

clean or in

disrepair.

□ No.

140

2.5* At least one toilet meets the

needs of people with reduced

mobility.

□ Yes. □ Yes, but not

available or in

disrepair.

□ No toilets for disabled

users.

2.6* Functioning hand hygiene

stations within 5 m of latrines.

□ Yes. □ Present, not

functioning or

no water or

soap.

□ Not present.

2.7* Record of cleaning visible and

signed by the cleaners each

day.

□ Yes. □ Toilets

cleaned but

not recorded.

□ No record / toilets

cleaned less than once a

day.

2.8* Wastewater is safely managed

through use of on-site

treatment (i.e. septic tank

followed by drainage pit) or

sent to a functioning sewer

system.

□ Yes. □ Present but

not

functioning.

□ Not present.

2.9* Greywater (i.e. rainwater or

wash water) drainage system

is in place that diverts water

away from the facility (i.e. no

standing water) and also

protects nearby households.

□ Yes. □ Yes, but not

functioning

and obvious

pools of water.

□ Not present.

2.10

*

Latrines are adequately lit,

including at night.

□ Yes □ Lighting

infrastructure

exists, but not

functioning.

□ Not adequately lit or

no lighting

infrastructure.

141

OVERALL FACILITY ASSESSMENT ENTER FACILITY CODE:_________________

BASELINE ASSESSMENT (Clearly tick the box that best describes

your findings) Assessment Date:

2.b HEALTHCARE WASTE Meets Target (+++)

Partially Meets Target

(++)

Does not Meet Target

(+)

Explain the reasoning for your scoring and comment as appropriate. For indicators with an asterisk, read

the notes before scoring.

2.11 A trained person is responsible for the management of healthcare waste in the healthcare facility.

□ Yes, presented and adequately trained.

□ Appointed but not trained.

□ Not appointed.

142

2.12*

Functional waste collection containers for 1) non-infectious (general) waste, 2) infectious waste and 3) sharps waste in close proximity to all waste generation points.

□ Yes. □ Separate bins present but lids missing or more than ¾ full; only two bins (instead of three); or at some but not all waste generation points.

□ No bins or separate sharps disposal.

2.13 Waste correctly segregated at all waste generation points.

□ Yes. □ Some sorting but not all correctly or not practiced throughout the facility.

□ No sorting.

143

2.14 Functional burial pit/fenced waste dump or municipal pick-up available for disposal of non-infectious (non-hazardous/general waste).

□ Yes. □ Pit in facility but insufficient dimensions; overfilled or not fenced and locked; irregular municipal waste pick up, etc.

□ No pit or other disposal method used.

2.15*

□ Incinerator or alternative treatment technology for the treatment of infectious and sharp waste is functional and of a sufficient capacity.

□ Yes. □ Present but not functional and/or of a sufficient capacity.

□ None present.

2.16 Sufficient energy available for incineration or alternative treatment technologies (mark if not applicable)

□ Yes, always.

□ Yes, sometimes.

□ Never.

144

2.17 Hazardous and non-hazardous waste are stored separately before being treated/disposed of or moved off site.

□ Yes, separated storage areas available.

□ Separated storage areas are available but with insufficient capacity or overfilled.

□ No separated storage areas available.

2.18*

All infectious waste is stored in a protected area before treatment, for no longer than the default and safe time.

□ Yes. □ Treated between 24-48 hours.

□ Treated after 48 hours or not treated at all.

2.19*

Anatomical- pathological waste is put in a dedicated pathological waste/placenta pit, burnt in a crematory or buried in a cemetery. (mark if not applicable).

□ Yes. □ Pit is present but not used or functional or overfilled or not fenced and locked.

□ None present.

145

2.20*

Dedicated ash pits available for disposal of incineration ash (mark if not applicable).

□ Yes. □ Present but not functional or overfilled or not fenced and locked.

□ None present.

2.21 Protocol or SOP (Standard Operating Procedure) for safe management of healthcare waste clearly visible and legible.

□ Yes, visible and implemented

□ Written but not visible or implemented.

□ No protocol/SOP in place.

2.22*

Appropriate protective equipment for all staff in charge of waste treatment and disposal.

□ Yes. □ Some equipment available, but not for all staff, or available but damaged.

□ None available.

146

OVERALL FACILITY ASSESSMENT

ENTER FACILITY

CODE:_________________

BASELINE ASSESSMENT

(Clearly tick the box that best describes your findings) Assessment Date:

3.a HAND HYGIENE

Meets

Target

(+++)

Partially Meets

Target

(++)

Does not

Meet Target

(+)

Explain the reasoning for your scoring

and comment as appropriate. For

indicators with an asterisk, read the notes

before scoring.

3.1* Functioning hand hygiene

stations are available at

all points of care.

□ Yes. □ Stations

present, but no

water and/or

soap or alcohol

hand-rub

solution.

□ Not present.

3.2* Hand hygiene promotion

materials clearly visible

and understandable at

key places.

□ Yes. □ Some places

but not all.

□ None.

3.3* Functioning hand hygiene

stations are available in-

service areas.

□ Yes. □ Stations

present, but no

water and/or

soap or alcohol

hand-rub

solution.

□ Not present.

147

3.4* Functioning hand hygiene

stations available in

waste disposal area

□ Yes. □ Stations

present, but no

water and/or

soap.

□ Not present.

3.5 Hand hygiene compliance

activities are undertaken

regularly.

□ Yes. □ Compliance

activities in

policy, but not

carried out with

any regularity.

□ No compliance

activities.

OVERALL FACILITY

ASSESSMENT

ENTER FACILITY

CODE:_________________

BASELINE ASSESSMENT

(Clearly tick the box that best describes your findings) Assessment Date:

3.b ENVIRONMENT,

CLEANLINESS AND

DISINFECTION

Meets

Target

(+++)

Partially Meets

Target

(++)

Does not

Meet Target

(+)

Explain the reasoning for your scoring

and comment as appropriate. For

indicators with an asterisk, read the

notes before scoring.

148

3.6 The exterior of the

facility is well-fenced,

kept generally clean (free

from solid waste,

stagnant water, no

animal and human

faeces in or around the

facility premises, etc.).

□ Yes. □ Partly but

improvements

could be made.

Yes,

sometimes.

□ Not kept clean at

all.

3.7 General lighting

sufficiently powered and

adequate to ensure safe

provision of healthcare

including at night (mark

if not applicable).

□ Yes, always. □ Yes,

sometimes.

□ Never.

3.8* Floors and horizontal

work surfaces appear

clean.

□ Yes. □ Some floors

and work

surfaces

appear clean,

but others do

not.

□ Most and/or all

floors and surfaces

are visibly dirty.

3.9 Appropriate and well-

maintained materials for

cleaning (i.e. detergent,

mops, buckets, etc.) are

available.

□ Yes. □ Yes, available

but not well

maintained.

□ No materials

available.

149

3.10* At least two pairs of

household cleaning

gloves and one pair of

overalls or apron and

boots in a good state, for

each cleaning and waste

disposal staff member.

□ Yes. □ Available but

in poor

condition.

□ Not available.

3.11 At least one member of

staff can demonstrate

the correct procedures

for cleaning and

disinfection and apply

them as required to

maintain clean and safe

rooms.

□ Yes. □ Procedure is

known but not

applied.

□ Procedure not

known or applied.

3.12 Beds have insecticide

treated nets to protect

patients from mosquito-

borne diseases.

□ Yes, on all

beds.

□ Available on

some but not

all beds, or

available but

with rips and or

holes.

□ No bed nets

available.

150

3.13 A mechanism exists to

track supply of IPC-

related materials (such

as gloves and protective

equipment) to identify

stock-outs.

□ Yes. □ Mechanism

exists but is not

enforced.

□ No mechanism

exists.

3.14 Record of cleaning visible

and signed by the

cleaners each day.

□ Yes. □ Record exists

but is not

completed

daily or is

outdated.

□ No record of

floors and surfaces

being cleaned.

3.15 Laundry facilities are

available to wash linen

from patient beds

between each patient.

□ Yes. □ Facilities

exist but are

not working or

not being used.

□ No facilities,

and/or no linen.

3.16 The facility has sufficient

natural ventilation and

where the climate

allows, large opening

windows, skylights and

other vents to optimize

natural ventilation.

□ Yes. □ Some

ventilation but

not well

maintained or

insufficient to

produce

natural

ventilation.

□ No.

151

3.17 Kitchen stores and

prepared food is

protected from flies,

other insects or rats.

□ Yes.

N/A

□ No.

3.18 Beds for patients should

be separated by a

distance of 2.5 metres

from the centre of one

bed to the other and

each bed has only one

patient.

□ Yes, all beds

meet this

guidance.

□ Some but not

all beds fit this

criteria.

□ No beds meet

this criteria.

OVERALL FACILITY ASSESSMENT

ENTER FACILITY

CODE:_________________

BASELINE ASSESSMENT

(Clearly tick the box that best describes your

findings)

Assessment Date:

4 MANAGEMENT

Meets

Target

(+++)

Partially Meets

Target

(++)

Does not

Meet Target

(+)

Explain the reasoning for your scoring

and comment as appropriate. For

indicators with an asterisk, read the

notes before scoring.

152

4.1 WASH FIT or other quality

improvement/management

plan for the facility is in

place, implemented and

regularly monitored.

□ Yes.

□ Complete but

has not been

implemented

and/or is not

monitored, or

incomplete.

□ No plan.

4.2* An annual planned budget

for the facility is available

and includes funding for

WASH infrastructure,

services, personnel and the

continuous procurement of

WASH items (hand hygiene

products, minor supplies to

repair pipes, toilets, etc.)

which is enough to meet the

needs of the facility.

□ Yes. □ Yes, but

budget is

insufficient.

□ No budget.

4.3 An up-to-date diagram of the

facility management

structure is clearly visible

and legible.

□ Yes. □ Yes, but not

up to date.

□ Not available.

153

4.4 Adequate cleaners and

WASH maintenance staff are

available.

□ Yes. □ Some

available, but

not adequate

or not skilled/

motivated.

□ None available.

4.5 A protocol for operation and

maintenance, including

procurement of WASH

supplies is visible, legible and

implemented.

□ Yes. □ Protocol

exists but not

implemented.

□ No protocol.

4.6* Regular ward-based audits

are undertaken to assess the

availability of hand-rub,

soap, single use towels and

other hand hygiene

resources.

□ Yes. □ Undertaken

less than once

a week or

assessment is

incomplete.

□ Not undertaken.

4.7 New healthcare personnel

receive IPC training as part of

their orientation program.

□ Yes. □ Some but not

all staff.

□ No training.

4.8 Healthcare staff are trained

on WASH/IPC each year.

□ Yes. □ Staff are

trained but not

every year or

only some staff

are trained.

□ No training.

154

4.9 Facility has a dedicated

WASH or IPC focal person.

□ Yes. □ Yes, but focal

point does not

have sufficient

time, resources

or motivation

to carry out

duties.

□ No.

4.10 All staff have a job

description written clearly

and legibly, including WASH-

related responsibilities and

are regularly appraised on

their performance.

□ Yes. □ Some, but

not all, staff

have a job

description, or

their

performance is

not appraised.

□ No job

description

written.

4.11 High performing staff are

recognized and rewarded

and those that do not

perform are dealt with

accordingly.

□ Yes. □ Either high or

low performers

addressed but

not both.

□ No action or

recognition of staff

based on

performance.

155

Sanitary Inspection Form - Notes

Do your assessment with the public health officer on site. Always write the name of the company supplying water to the facility

(where applicable).

ENTER FACILITY CODE: _____________________________

156

SANITARY INSPECTION FORM 1: DUG WELL WITH HAND PUMP

I. General information

a. Code of facility: b. Location and/or name of dug well: c. Date of inspection: d. Weather conditions during inspection:

Note. If there is more than one dug well accessed by the facility, or if the facility uses other water sources (such as springs or boreholes), carry out sanitary inspections for these sources too.

II. Specific questions for assessment

1. Is the source located at an unsafe distance from an unsealed latrine (i.e. a latrine near is uphill or at a location where the groundwater gradient would flow from the latrine to the water source)? Y/N

2. Is the fence absent, inadequate or faulty? Y/N 3. Can animals have access within 30 metres (100 feet) of the well? Y/N 4. Is there any other source of pollution within 30 metres (100 feet) of the

well (such as animal breeding, farming, roads, healthcare waste, domestic garbage)? Y/N

5. Is there stagnant water within 3 metres of the well? Y/N 6. Is the drainage channel absent or cracked, broken or in need of

cleaning? Y/N 7. Is the cement floor or slab less than 2 metres in diameter around the

top of the well? Y/N 8. Are there cracks in the cement floor or slab? Y/N 9. Is the hand pump loose at the point of attachment or, for rope-washer

pumps, is the pump cover missing or damaged? Y/N

10. Is the well cover absent, cracked or insanitary? Y/N

Total score of risk factors as total number of “YES” answers: III. Results and comments

a. Sanitary inspection risk score (tick appropriate box):

Very high risk (risk score: 9–10)

High risk (risk score: 6–8)

Medium risk (risk score: 3–5)

Low risk (risk score: 0–2)

157

b. Important points of risk noted and reported on the reverse of this form:

• list according to question numbers 1–10

• additional comments.

IV. Names and signatures of assessors: ………………………………………………………………. SANITARY INSPECTION FORM 2: DEEP BOREHOLE WITH MOTORIZED PUMP

I. General information

a. Code of facility: b. Location and/or name of borehole: c. Date of inspection: d. Weather conditions during inspection:

Note. If there is more than one borehole accessed by the facility, or if the facility uses other water sources (such as springs or dug wells), carry out sanitary inspections for these sources too.

II. Specific questions for assessment 1. Is there a latrine or sewer within 15–20 m of the extraction site/well-head? Y/N 2. Is the nearest latrine a pit latrine that percolates to soil, i.e. not connected to a septic tank or sewer? Y/N 3. Is there any other source of pollution (e.g. animal excreta, rubbish, surface water) within 10 m of the borehole? Y/N 4. Is there an uncapped well within 15–20 m of the borehole? Y/N 5. Is the drainage area around the pump house faulty? Y/N 6. Is the fencing around the installation damaged in any way which would permit any unauthorized entry or allow animals access? Y/N 7. Is the floor of the pump house permeable to water? Y/N 8. Is the well seal unsanitary? Y/N 9. Is the chlorination functioning properly? Y/N

158

10. Is chlorine present at the sampling tap? Y/N Total score of risk factors as total number of “YES” answers: III. Results and comments

a. Sanitary inspection risk score (tick appropriate box):

Very high risk (risk score: 9–10)

High risk (risk score: 6–8)

Medium risk (risk score: 3–5)

Low risk (risk score: 0–2)

b. Important points of risk noted and reported on the reverse of this form:

• list according to question numbers 1–10

• additional comments.

IV. Names and signatures of assessors: ………………………………………………………………. SANITARY INSPECTION FORM 3: PUBLIC/YARD TAPS AND PIPED DISTRIBUTION I. General information

a. Code of facility: b. Date of inspection: c. Weather conditions during inspection: d. Location and/or name of water source(s) feeding the distribution system: e. Location and/or name of storage reservoir feeding the distribution system (if

any): Note. If the distribution system is served by a storage reservoir, also carry out a sanitary inspection using the form “Storage reservoirs”.

II. Specific questions for assessment

Note. Fill in one form per public or yard tap under inspection. In facilities with water piped directly into the building only questions 7–10 apply. Not all taps within the facility need to be inspected in every inspection round; a selected sample is enough.

Public or yard tap 1. Does the tap leak? Y/N 2. Is the tap or are attachments (such as hoses) insanitary? Y/N 3. Does spilt water accumulate around the tap stand? Y/N 4. Is the area around the tap stand polluted by waste, faeces or other

materials? Y/N

159

5. Is the area around the tap stand unfenced, allowing animals to access the area? Y/N

6. Is there a sewer or a latrine at an unsafe distance from the tap stand (generally 30 m but may be more or less depending on the gradient, geology and size of water or sewer infrastructure) Y/N

Piped distribution 7. Are there any signs of leaks in the inspection area (for example,

accumulating water)? Y/N 8. Are any of the pipes exposed above ground in the inspection area? Y/N 9. Have users report any pipe breaks within the last week? Y/N 10. Has there been discontinuity in the last 10 days? Y/N

Total score of risk factors as total number of “YES” answers: III. Results and comments

a. Sanitary inspection risk score (tick appropriate box): Note. In situations where, only questions 6–10 apply, the score below can be adapted as follows: “Very high” = 5; “High” = 3–4; “Medium” = 2; “Low” = 0–1.

Very high risk (risk score: 9–10)

High risk (risk score: 6–8)

Medium risk (risk score: 3–5)

Low risk (risk score: 0–2)

b. Important points of risk noted and reported on the reverse of this form:

• list according to question numbers 1–10

• additional comments.

IV. Names and signatures of assessors: ………………………………………………………………. SANITARY INSPECTION FORM 4: RAINWATER HARVESTING I. General information

a. Code of facility: b. Location and/or name of rainwater storage: c. Date of inspection: d. Weather conditions during inspection:

Note. If the facility uses other water sources (such as springs or boreholes), carry out sanitary inspections for these sources too.

II. Specific questions for assessment

160

1. Is there any visible contamination of the roof catchment area (plants, dirt, or excreta)? Y/N 2. Are the guttering channels that collect water dirty? Y/N 3. Is there any deficiency in the filter box at the tank inlet (e.g. lacks fine gravel)? Y/N 4. Is there any other point of entry to the tank that is not properly covered? Y/N 5. Is there any defect in the walls or top of the tank (e.g. cracks) that could let water in? Y/N 6. Is the tap leaking or otherwise defective? Y/N 7. Is the concrete floor under the tap defective or dirty? Y/N 8. Is the water collection area inadequately drained? Y/N 9. Is there any source of pollution around the tank or water collection area (e.g. excreta)? Y/N 10. Is a bucket in use and left in a place where it may become contaminated? Y/N Total score of risk factors as total number of “YES” answers: III. Results and comments

a. Sanitary inspection risk score (tick appropriate box):

Very high risk (risk score: 9–10)

High risk (risk score: 6–8)

Medium risk (risk score: 3–5)

Low risk (risk score: 0–2)

b. Important points of risk noted and reported on the reverse of this form:

• list according to question numbers 1–10

• additional comments. IV. Names and signatures of assessors: ……………………………………………………………….

161

SANITARY INSPECTION FORM 5: STORAGE RESERVOIRS I. General information

a. Code of facility: b. Location and/or name of storage reservoir: c. Date of inspection: d. Weather conditions during inspection: e. Location and/or name of water source(s) feeding the reservoir:

Note. If there is more than one storage reservoir used in your facility, use one form for each reservoir. Note. If the storage reservoir feeds a piped distribution system, also carry out a sanitary inspection using the form “Public/yard taps and piped distribution”. Note. If the storage reservoir is equipped with a tap for collecting water, also carry out a sanitary inspection using questions 1–5 of the form “Public/yard taps and piped distribution”.

II. Specific questions for assessment

1. Is there any point of leakage of the pipe between source and storage reservoir? Y/N

2. Is the physical infrastructure of the storage reservoir cracked or leaking? Y/N

3. Is the inspection cover of the storage reservoir absent or open? Y/N 4. Is the inspection cover faulty, corroded or is the concrete around the

cover damaged? Y/N 5. Is the inspection cover visibly dirty? Y/N 6. Are screens protecting the air vents on the storage reservoir missing or

damaged? Y/N 7. If there is an overflow pipe, is the screen protecting it missing or

damaged? Y/N 8. Is there any scum or foreign object in the storage reservoir? Y/N 9. Is the diversion ditch above the storage reservoir absent or non-

functional? Y/N 10. Is the area around the storage reservoir unfenced or is the fence

damaged, allowing animals to access the area? Y/N 11. Is the storage reservoir not regularly cleaned and disinfected? Y/N

Total score of risk factors as total number of “YES” answers:

162

III. Results and comments

a. Sanitary inspection risk score (tick appropriate box):

Very high risk (risk score: 9–11)

High risk (risk score: 6–8)

Medium risk (risk score: 3–5)

Low risk (risk score: 0–2)

b. Important points of risk noted and reported on the reverse of this form:

• list according to question numbers 1–11

• additional comments.

IV. Names and signatures of assessors:…………………………………………………………..

163

S3 File: Consolidated criteria for reporting qualitative studies (COREQ)

Manuscript: Evaluating the foundations that help avert antimicrobial resistance:

Performance of essential water sanitation and hygiene functions in hospitals

and requirements for action in Kenya

Consolidated criteria for reporting qualitative studies (COREQ): 32-item

checklist

Developed from:

Tong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups. International Journal for Quality in Health Care. 2007. Volume 19, Number 6: pp. 349 – 357

No. Item

Guide questions/description Reported on

Page #

Domain 1: Research

team and reflexivity

Personal Characteristics

1. Interviewer/facilitator Which author/s conducted the

interview or focus group?

9

2. Credentials What were the researcher’s

credentials? E.g. PhD, MD

1

3. Occupation What was their occupation at the time

of the study?

9

4. Gender Was the researcher male or female? 1

5. Experience and

training

What experience or training did the

researcher have?

9

Relationship with

participants

6. Relationship

established

Was a relationship established prior to

study commencement?

Yes, 9

7. Participant knowledge

of the interviewer

What did the participants know about

the researcher? e.g. personal goals,

reasons for doing the research

9

164

8. Interviewer

characteristics

What characteristics were reported

about the interviewer/facilitator? e.g.

Bias, assumptions, reasons and

interests in the research topic

9

Domain 2: study design

Theoretical framework

9. Methodological

orientation and Theory

What methodological orientation was

stated to underpin the study? e.g.

grounded theory, discourse analysis,

ethnography, phenomenology,

content analysis

9

Participant selection

10. Sampling How were participants selected? e.g.

purposive, convenience, consecutive,

snowball

10

11. Method of approach How were participants approached?

e.g. face-to-face, telephone, mail,

email

8/9

12. Sample size How many participants were in the

study?

8

13. Non-participation How many people refused to

participate or dropped out? Reasons?

None, 9

Setting

14. Setting of data

collection

Where was the data collected? e.g.

home, clinic, workplace

9

15. Presence of non-

participants

Was anyone else present besides the

participants and researchers?

9

16. Description of sample What are the important characteristics

of the sample? e.g. demographic data,

date

8

Data collection

17. Interview guide Were questions, prompts, guides

provided by the authors? Was it pilot

tested?

8/9

18. Repeat interviews Were repeat interviews carried out? If

yes, how many?

No, 9

19. Audio/visual

recording

Did the research use audio or visual

recording to collect the data?

Yes, 9

165

20. Field notes Were field notes made during and/or

after the interview or focus group?

No

21. Duration What was the duration of the

interviews or focus group?

9

22. Data saturation Was data saturation discussed? Yes, 9

23. Transcripts returned Were transcripts returned to

participants for comment and/or

correction?

No,– 10

Domain 3: analysis and

findings

Data analysis

24. Number of data

coders

How many data coders coded the

data?

10

25. Description of the

coding tree

Did authors provide a description of

the coding tree?

Yes 10

26. Derivation of themes Were themes identified in advance or

derived from the data?

Yes, from

Data

27. Software What software, if applicable, was used

to manage the data?

10

28. Participant checking Did participants provide feedback on

the findings?

Yes, 9

Reporting

29. Quotations

presented

Were participant quotations presented

to illustrate the themes/findings? Was

each quotation identified? e.g.

participant number

Yes

30. Data and findings

consistent

Was there consistency between the

data presented and the findings?

Yes

31. Clarity of major

themes

Were major themes clearly presented

in the findings?

Yes

32. Clarity of minor

themes

Is there a description of diverse cases

or discussion of minor themes?

From page 17-

24

166

S 4 File. Ward level Aggregate by Domain for Water, Hygiene and Organisation Management

Multi-panel plot showing the mean service performance at ward level for all the water, hygiene and organisational management domain.

167

S5 File. Performance of Indicators under the infection prevention and control committee and

hospital management

Shows mean service performance at ward level for indicators under the infection prevention and

control and hospital management domains shown by the vertical bars. The horizontal bars

summarise the performance of each indicator across all the hospitals. The central tiles in the

central grid are coloured according to the performance classification of each indicator in each

hospital using % cut-offs.

168

169

Chapter 4

Infection prevention and control during

the COVID-19 pandemic: challenges and

opportunities for Kenyan public hospital

Michuki Maina, Olga Tosas-Auguet, Mike English, Constance

Schultsz and Jacob McKnight [version 1; peer review: 2

approved]. Wellcome Open Res 2020, 5:211

(https://doi.org/10.12688/wellcomeopenres.16222.1)

170

Abstract

Background

Infection prevention and control, and water sanitation and hygiene have an

essential role in ensuring the quality of care and patient outcomes in hospitals.

Using a modification of the World Health Organization’s water sanitation and

hygiene facility improvement tool, we undertook assessments in 14 public

hospitals in Kenya in 2018. The hospitals received written feedback on areas

where they could make improvements. Following the first confirmed cases of

COVID-19 in Kenya, we were drawn to ask whether the results of our pre-

pandemic survey had led to action, and whether or not the threat of COVID-19

had focused more attention on infection prevention and control and water

sanitation and hygiene.

Methods

Using a semi-structured interview guide, we carried out phone interviews with

key hospital leaders in 11 of the 14 hospitals. The data were transcribed and

coded into thematic areas. We draw on these interviews to describe the status

and awareness of infection prevention and control.

Results

The infection prevention and control committee members are training health

workers on infection prevention and control procedures and proper use of

personal protective equipment and in addition, providing technical support to

hospital managers. While some hospitals have also accessed additional funds

to improve infection prevention and control, they tended to be small amounts

of money. Long-standing challenges with supplies of infection prevention and

control materials and low staff morale persist. Crucially, the reduced supply of

171

personal protective equipment has led to fear and anxiety among health care

personnel.

Conclusions

As funds are mobilised to support care for COVID-19, we ask that funds

prioritise infection prevention and control measures. This would have a

profoundly positive effect on within hospital virus transmission, patient and

staff safety but also lasting benefits beyond the COVID-19 pandemic.

Key Words

Water Sanitation and Hygiene, Infection Prevention and Control, COVID- 19

172

Introduction

Improving water sanitation and hygiene (WASH) and infection prevention and

control (IPC) in medical facilities is crucial to the fight against COVID-19 in all

countries, but particularly so in those countries with long-standing problems in

maintaining effective infection control (1). Hospitals are now thought to have

been significant sites for COVID-19 infection spread in some of the countries

worst affected by the pandemic (2). Ensuring effective infection control in

hospitals protects patients and health workers, improving worker confidence

and morale. Maintaining a healthy workforce is essential to the COVID-19

response and the provision of all other health interventions(3).

We conducted a WASH survey of 14 Kenyan hospitals in 2018. We adapted the

World Health Organisation's WASH facility improvement tool to explore

performance differences between hospitals and between wards in individual

hospitals (4). We found a wide degree of performance variation and particular

problems in hand hygiene and waste management infrastructure. We also

proposed accountability structures to facilitate improvements in WASH (5, 6).

All the 14 original hospitals had an IPC focal person who was trained to conduct

WASH-FAST surveys and had surveyed at least four hospitals, including their

own; each hospital received detailed site-specific feedback on their results. We

hoped that this feedback and the training given to local focal points might

prompt local post-survey WASH and IPC improvement efforts.

The first confirmed case of COVID-19 in Kenya was announced on 12th March

2020. This was accompanied by extensive media coverage and citizen education

on COVID-19, including the importance of social distancing and proper hand

hygiene. At this point, the main policy measures implemented by the

government included quarantine of travellers returning to Kenya, the

suspension of public gatherings and the setting up of a national COVID-19

173

isolation unit in Nairobi (7). No restriction of movement or curfews were

executed at this point. Following the first confirmed cases, we were drawn to

ask what actions had occurred, and whether or not the threat of COVID-19 was

focusing more attention on IPC and WASH. On 2nd April 2020, we contacted the

hospital staff who previously led the WASH surveys. By this date, there were

110 confirmed cases around the country, and the main government

interventions included restriction of movement and a nationwide dusk to dawn

curfew. The government, through the regional governments, had embarked on

setting up COVID-19 isolation units in multiple county hospitals (8). As April

proceeded, Kenya requested World Bank funding for these isolation units, for

training and for the purchase of ventilators to improve intensive care capacity

(9).

Methods

Interview data collection

We conducted semi-structured phone interviews in April and May 2020. Each

interview lasted approximately 20-30 minutes, with the consent of 11

representatives of the original 14 hospitals, (the semi-structured interview

template can be found in the online data repository). The respondents were

medical officers, dentists, pharmacists, non-physician clinicians, nurses, and

public health officers. Of these, one was a hospital director, and another was in

charge of hospitals across the administrative region (sub-county). The

interviewees comprised a balanced mix of gender, age and experience. The

semi-structured interview instrument used was informed by data collected

from the previous surveys. The interview guide was prepared by all the authors

who consist of a mix of clinicians, epidemiologists, health system experts and

medical anthropologists. The authors are also very familiar with the Kenyan

health system and IPC context and were all actively involved in the previous

174

IPC/WASH surveys in 2018 (5, 6). The first author conducted the phone

interviews. The first author, who is a Kenya doctor, also led the original surveys

and is familiar with the health system context. Field notes were not taken.

Background information concerning the study and interview was availed to

these interviewees before consenting. All respondents gave verbal consent for

the interviews.

Analysis of data obtained from interviews

We sought to understand if any changes or improvements had occurred within

the hospital for each area we had previously analysed (5, 6). We then asked

how the present COVID-19 pandemic was affecting efforts to improve the

WASH infrastructure and IPC, how the facility was preparing for COVID- 19

patients and how the pandemic had affected hospital and health worker

activities. The audio files of each interview were transcribed and uploaded into

NVivo 12 and kept on an encrypted laptop. The first author coded the

transcripts independently before discussing the codes and agreeing on

combined axial codes and thematic areas with the last author. The findings from

these themes illustrated using quotes. No repeat interviews were conducted. A

Consolidated criteria for Reporting Qualitative research (COREQ) checklist was

completed and is included in the data repository.

Ethical approval

This study received approval from the Oxford Tropical research ethics

committee (OXTREC) from the University of Oxford (Ref: 525–17) and the

Kenyan Medical Research Institute (Ref: KEMRI/SERU/CGMR-C//086/3450).

175

Results

Infection control status and awareness improved

Our original survey had noted the low performance of WASH and infection

control. Previously, staff responsible for hygiene issues contrasted infection

control committees with those of more prestigious domains, such as the

medicines committee, and complained that members did not attend meetings.

Since the emergence of COVD-19, this has changed and the roles played by the

IPC committees have gained more focus and importance.

"Actually, right now people are more responsive, previously

you call people on things to do with IPC and all that people

used to take it for public health workers and all that. But now

we are even having the consultants attending."

IPC Committee Member

IPC committees are also now being asked to offer technical advice to managers.

"But now we also do have supervision. The committee does

the supervision within the hospital, and then reports whatever

we have agreed and seen to the CEO."

Hospital IPC coordinator

Hence, infection control has become an urgent priority, and the interviews

revealed hospital staff at all levels are now concerned about IPC best practice.

This may be due to the fear of contracting COVID-19 at work, and it was clear

that these concerns are leading to a great deal of anxiety:

"There is a lot of fear initially the time we were starting out.

And after the once who have gone through things change for

176

example initially, they could say they can't see patients

without an N95 mask, but after the training on which masks to

be worn where people become more confident. The element of

fear and anxiety is still there, but the ones who attend the

training it becomes less." IPC Leader

Positive changes

The interest in preparing for COVID-19 has resulted in positive actions in several

areas: In some of the facilities, this has offered an opportunity to increase the

IPC supplies and infrastructure including hand hygiene facilities in the hospital,

albeit some of these are temporary.

"I believe that after this we will go far, we never had sanitisers

before, these days we have sanitiser." Hospital IPC leader.

Where previously budgets for IPC activities were non-existent, in some

hospitals some funds are now available specifically for IPC activities.

"I don't know where the med sup [Medical Director] got some

little money. Almost USD 1500, he was telling us to come up

with a budget. We sat last week. And we have given the

proposal to him." IPC Committee Leader.

Some of the facilities reported an improvement in the cleanliness standards.

This was mainly through reallocation of cleaning staff and increased supervision

in critical areas of the hospitals to ensure these were regularly cleaned, and

waste management was improved.

"And then the cleanliness, we improved on the supervision.

Yeah, we got another public health officer. And we got at least

177

two supervisors for the cleaning company. And then with the

supervision, there is some, there is an improvement." IPC Lead.

Some of the facilities reported significant improvement in WASH infrastructure

including fixing of hand hygiene stations;

"The sinks even if they are not fully functional, maybe we are

over 90% there. Because at least you can repair one and it

gets spoilt. But I'm telling you we are very far. That one we

have done it." IPC Lead.

Ongoing IPC challenges

While the respondents noted areas of improvement, the overall assessment of

our respondents suggested that long-standing issues had not been adequately

resolved. These include challenges with waste management, infrastructure and

funding. Some of the areas where challenges persisted included in

infrastructure where, for example, poor quality fixtures of hand hygiene (taps

and pipes) led to frequent breakages and leaks. Other areas with long-standing

challenges were waste management and funding for IPC activities. All these

play a critical role in service delivery and staff morale.

"The other challenge is that currently, our incinerator is down.

So, we are not able to offer incineration services. Both in

house and even anyone from outside. So, what we are doing

right now we are disposing our waste off sight."

IPC Committee Leader.

"But of course, here, staff morale has gone down because of

the issue of salaries – it's so delayed. Like now, we have not

even got our salaries you can imagine." Hospital Manager.

178

Newer challenges have also emerged. Though some of the IPC supplies are

present, in some facilities, the supply diminishes rapidly due to theft and

misuse.

"Even the soap, you put it outside there, and the following day

even the container is not there. Yes, sometimes we put the

soap there and the relatives of the patients when they come

in, they think that maybe that is the soap they should be using

for washing clothes. They drain it all out in the basins when

they come across it. They are taking it away." IPC lead.

Health worker safety

In addition to these long-standing issues, a host of new issues related to the

pandemic response have been raised, and most prominent among these are

concerns about the safety of health workers. These include challenges in the

acquisition of personal protective equipment (PPE) for both clinical and support

staff.

"Right now, it's not adequate because we are yet to get the

gowns, the masks and the gloves we have. But the gowns

which I think we really need we don't have. But at the

designated areas they have some." IPC Leader.

The lack of PPE is strongly related to levels of anxiety and concerns about the

ability of the workforce to do its job:

"Yes, we are very, very anxious. We are so worried about

ourselves, our safety before even the patients because we

have families definitely. So, there is this sort of some attitude."

Doctor.

179

"Everybody is worried. Everybody, but if they provide people

with enough PPE'S, nobody is there to worry. People are ready

to work." IPC Leader.

Health seeking behaviour in the COVID-19 pandemic

In terms of patients seeking care for other ailments, we noted the numbers

have significantly reduced in the facilities. However, the hospitals that serve

large populations still have to contend with crowding. This is a challenge in

reducing hospital-acquired infections.

"The management had decided to halt outpatient clinics;

unfortunately, we still have a problem with that. like today is

our Gynaecology Outpatient Clinic, so we expected we

wouldn't have any patients, but apparently, we were so

disappointed when we found like about thirty patients and you

know they are all crammed together, there is no even social

distancing or anything" Medical Doctor

We contrast this to other facilities where numbers are worryingly low. The

concern is some of these patients will lack much-needed care. This may, in the

long run, affect the gains made in reducing the infant and maternal mortalities

in the region.

Key insights

Huge opportunity for improvements due to status change and interest.

The considerable importance of WASH and IPC in addressing COVID-19 provides

a real opportunity for improvement of IPC in health facilities. The existing

structure of IPC committees and coordinators, and public health officers, could

offer a real opportunity for highly efficient use of COVID-19 funds. A bottom-up

180

approach would ensure funds are allocated where they are needed. Some

hospitals have embraced this and are making positive changes, even with

limited resources.

"…I was given the responsibility of like a coordinator; I usually

go around on a daily basis. You decide today to go to the

medical department, maybe the following day you can go to

maternity, you can go to paediatrics. You look at the way they

are segregating you find that they are not doing it properly

you just correct them." IPC Lead.

Structural weaknesses, long-standing issues need to be prioritised and

dealt with urgently.

The issues of lack of crucial infrastructure still exist. These include poor waste

disposal, plumbing to guarantee water supply in the wards and proper

ventilation. Improving plumbing and access to appropriate sinks are some of

the interventions that can be effected quickly. Others may include long term

investments like improving ventilation within the hospitals by the installation

of exhaust fans. Other long-term investments would consist of setting

up/building fully-fledged isolation units.

However, these investments need to be prioritised in budget allocations as they

are essential for the appropriate management of COVID-19 and other infectious

conditions which are prevalent in our region. Investing in IPC and WASH would

greatly complement the efforts being put to improve the availability of oxygen

and critical care support in some of the hospitals in our survey and across the

country.

181

Health systems challenges

The greatest risk COVID-19 poses in our context is a health system collapse. In

most facilities, though the patient load has significantly reduced, there is still a

need to ensure the health system does not get overwhelmed by COVID-19.

Health system collapse and closure of services such as maternal and child health

would result in tragic increases in maternal and child mortality (10). There have

been numerous reports of frontline health workers globally contracting COVID-

19 in the line of duty and some due to lack of proper PPE (11, 12). If health

workers do not feel safe while offering services, some may choose to stay away

or take industrial action, including strikes. Strikes have been frequent in Kenya

and the region with devastating effects, and there are ongoing negotiations to

avert a looming health worker strike in the country (13, 14).

Motivating staff and ensuring safe working conditions, including PPE, training

and treatment if they fall ill needs to be a priority at this time. With already low

staff numbers, staff staying off work due to illness or fear of COVID-19 as a

result of weak IPC structures would easily overwhelm service delivery within

hospitals. We can seize this moment to make things better not only during this

pandemic but for years to come.

"but again, COVID- 19 is a blessing in disguise it's not the

right phrase to use but I think this COVID coming, it's a

wakeup call to all these people who were like taking things for

granted. Our health system is in trouble; it has been in trouble

all through, now at least they are seeing." Medical Doctor.

Conclusion

Effective treatment of COVID-19 relies on effective IPC and WASH and funds

must be immediately directed towards this in low- and middle-income

182

countries such as Kenya where performance is often poor. These gains could

have long-lasting positive effects for other huge issues facing hospitals such as

antimicrobial resistance and hospital-acquired infections. We believe that

investing in IPC should be prioritised in advance of setting up critical care units

as part of the COVID-19 response (9). If we wish to mitigate the effects of

COVID-19 on health systems and avert the collapse of an already stretched

health system, IPC and WASH should be recognised as a major policy priority.

Staff safety should be prioritised by availing proper PPE for health workers in

their line of duty and during procedures that are aerosol-generating to ensure

they remain disease-free.

Availability of Data and Materials

All summary data, questionnaires and the Consolidated criteria for reporting

qualitative studies (COREQ) from the surveys are hosted in a public repository

Harvard Data verse. DOI Information: https://doi.org/10.7910/DVN/IJUWWR.

Applications for access of the study transcripts can be made through the Data

Governance Committee with details available on www.kemri-wellcome.org, or

email to cgmrc@kemri-wellcome.org.

Reporting guidelines

The Consolidated criteria for reporting qualitative studies (COREQ) was used for

this report.

Competing interests

There are no conflicts of interest declared by the authors.

Funding

This work was supported by funds from the economic and social research

council ESRCS # ES/P004938/1, and a Senior Research Fellowship awarded to

M.E. by The Wellcome Trust (207522). MM is supported by a grant from by the

183

Initiative to Develop African Research Leaders (IDeAL) through the DELTAS

Africa Initiative [DEL-15-003], an independent funding scheme of the African

Academy of Sciences (AAS)'s Alliance for Accelerating Excellence in Science in

Africa (AESA) and supported by the New Partnership for Africa's Development

Planning and Coordinating Agency (NEPAD Agency) with funding from the

Wellcome Trust [107769] and the U.K. government. The funders had no role in

drafting nor the decision for submitting this manuscript.

Acknowledgements

The authors would like to thank all the hospitals that participated in the initial

surveys and follow up interviews. This work is published with the permission of

the director of KEMRI.

184

References

1. World Health Organization. Water, sanitation, hygiene, and waste management for the

COVID-19 virus: interim guidance, 23 April 2020. World Health Organization; 2020.

2. Nacoti M, Ciocca A, Giupponi A, Brambillasca P, Lussana F, Pisano M, et al. At the

Epicenter of the Covid-19 Pandemic and Humanitarian Crises in Italy: Changing Perspectives on

Preparation and Mitigation. Catalyst non-issue content. 2020;1(2).

3. Ministry of Health. Interim Infection Prevention and Control Recommendations for

Coronavirus Disease 2019 (COVID-19) in Health Care Settings. Nairobi Ministry of Health 2020.

4. World Health Organization, UNICEF. Water and Sanitation for Health Facility

Improvement Tool (WASH FIT): a practical guide for improving quality of care through water,

sanitation and hygiene in health care facilities. Geneva; 2017. Report No.: 9241511699.

5. Maina M, Tosas-Auguet O, McKnight J, Zosi M, Kimemia G, Mwaniki P, et al. Extending

the use of the World Health Organisations' water sanitation and hygiene assessment tool for

surveys in hospitals - from WASH-FIT to WASH-FAST. PLoS One. 2019;14(12):e0226548.

6. Maina M, Tosas-Auguet O, McKnight J, Zosi M, Kimemia G, Mwaniki P, et al. Evaluating

the foundations that help avert antimicrobial resistance: Performance of essential water

sanitation and hygiene functions in hospitals and requirements for action in Kenya. PLoS One.

2019;14(10):e0222922.

7. First Case of coronavirus confirmed in Kenya [press release]. Nairob, 13th March 2020

2020.

8. National Emergency response committee on coronavirus update of coronavirus in the

country and response measures as at 2nd April 2020 [press release]. 2nd April 2020 2020.

9. Zhao Y, McKnight J, English M. Low-income countries’ bids for World Bank funding raise

serious concerns about their coronavirus strategies: The Conversation 2020 [Available from:

https://theconversation.com/low-income-countries-bids-for-world-bank-funding-raise-serious-

concerns-about-their-coronavirus-strategies-138628.

10. Roberton T, Carter ED, Chou VB, Stegmuller AR, Jackson BD, Tam Y, et al. Early estimates

of the indirect effects of the COVID-19 pandemic on maternal and child mortality in low-income

and middle-income countries: a modelling study. The Lancet Global Health.

11. Herron JBT, Hay-David AGC, Gilliam AD, Brennan PA. Personal protective equipment

and Covid 19- a risk to healthcare staff? The British journal of oral & maxillofacial surgery. 2020.

185

12. Centers for Disease Control and Protection. Characteristics of Health Care Personnel

with COVID-19 — United States, February 12–April 9, 2020 2020 [MMWR Morb Mortal Wkly Rep

2020;69:477–481:[

13. Adam MB, Muma S, Modi JA, Steere M, Cook N, Ellis W, et al. Paediatric and obstetric

outcomes at a faith-based hospital during the 100-day public sector physician strike in Kenya.

BMJ Global Health. 2018;3(2):e000665.

14. Magdalne Saya. Crisis looms as health care workers issue 7-day strike notice. The Star

2020 23rd April 2020.

186

187

Addendum 1

COVID-19: an opportunity to improve infection

prevention and control in low-income and

middle-income countries

Michuki Maina, Olga Tosas- Auguet, Mike English, Constance Schultsz

and Jacob McKnight

The Lancet Global Health, Volume 8, Issue 10, e1261

https://doi.org/10.1016/S2214-109X(20)30352-1

188

CORRESPONDENCE

We congratulate Paul Sonenthal and colleagues (1), for the timely assessment

of COVID-19 preparedness in Malawi. The authors highlight clear gaps in

infection prevention and control, including the availability of personal

protective equipment, but cautiously suggest that these findings might not be

generalisable to other low-income and middle-income countries. We did

detailed surveys before the COVID-19 pandemic to evaluate water sanitation

and hygiene, and infection prevention and control preparedness at 14 Kenyan

public hospitals(2). From the surveys, we noted that, even for these large

facilities, there were challenges in providing adequately treated water.

Additionally, there were scarce resources available to install new hand hygiene

and waste disposal structures, especially in locations where they were absent

(2).

We found that the responsibility for water sanitation and hygiene, and infection

prevention and control in hospitals is often poorly defined. Working with

stakeholders, we developed a framework to assign responsibility to specific

groups within the hospital and regional governments on the basis of their

capacity for action. This framework allowed us to identify the specific groups

who were best placed to tackle the problems we had highlighted. One of the

important groups identified was the hospital's infection prevention and control

committee(3). Our pre-pandemic survey suggested that these committees

were inactive and poorly constituted(3). The low status and awareness of

infection prevention and control in the hospitals contributed to these

committees being largely dormant(2). However, from our interviews with

health-care workers done during the COVID-19 pandemic in April 2020, we

found that these infection prevention and control committees are playing a

189

crucial role in training medical staff on personal protective equipment use and

infection prevention and control procedures (4).

These committees can provide strategic leadership on the purchase of good

quality infection prevention and control materials and address any deficiencies

in the hand hygiene and waste management infrastructure. With infection

prevention and control playing a crucial role in the management of COVID-19,

it provides an opportunity for the proper training of all health workers on the

best infection prevention and control practices (4).

The COVID-19 pandemic provides an opportunity to strengthen crucial aspects

of the health-care system that have been previously overlooked (5). These

aspects include the infection prevention and control infrastructure and the

formation of active infection prevention and control committees to provide

much-needed leadership.

These improvements will prove to be valuable not only during the COVID-19

pandemic, but also in the fight against other infectious diseases and

antimicrobial resistance.

We declare no conflict of interest.

190

References

1. Sonenthal PD, Masiye J, Kasomekera N, et al. COVID-19 preparedness in Malawi: a

national facility-based critical care assessment. The Lancet Global Health 2020; 8(7): e890-e2.

2. Maina M, Tosas-Auguet O, McKnight J, et al. Evaluating the foundations that help avert

antimicrobial resistance: Performance of essential water sanitation and hygiene functions in

hospitals and requirements for action in Kenya. PLoS One 2019; 14(10): e0222922.

3. Maina M, Tosas-Auguet O, McKnight J, et al. Extending the use of the World Health

Organisations' water sanitation and hygiene assessment tool for surveys in hospitals - from

WASH-FIT to WASH-FAST. PLoS One 2019; 14(12): e0226548.

4. Maina M, Tosas-Auguet O, English M, Schultsz C, McKnight J. Infection prevention and

control during the COVID-19 pandemic: challenges and opportunities for Kenyan public hospitals.

Wellcome open research 2020; 5: 211.

5. Zhao Y MJ, English M. . Low-income countries’ bids for World Bank funding raise serious

concerns about their coronavirus strategies. 18th May 2020 2020.

https://theconversation.com/low-income-countries-bids-for-world-bank-funding-raise-serious-

concerns-about-their-coronavirus-strategies-138628.

191

Addendum 2

Performance in water sanitation and hygiene

and infection prevention and control in Kenyan

hospitals and relevance for the COVID- 19

pandemic

Policy Brief, KEMRI Wellcome Trust Research Programme

October 2020

192

Key Messages

PERFORMANCE IN WATER SANITATION AND HYGIENE AND INFECTION

PREVENTION AND CONTROL IN KENYAN HOSPITALS AND RELEVANCE

FOR THE COVID-19 PANDEMIC

1. Poor water sanitation and hygiene (WASH) and infection prevention

and control (IPC) in health facilities increase the risk of hospital-

acquired infections including COVID-19 among patients, health care

workers and the community which results in greater use of second-line

antibiotics and drives antimicrobial resistance.

1. There are significant differences in the state of WASH within hospitals

wards and across hospitals. The aggregate hospital scores ranged

between 47% and 71%.

2. With the emergence of the COVID- 19 pandemic in Kenya, the IPC

committees which were previously dormant, have stepped up to offer

leadership and technical support to hospital managers.

3. The overall status of IPC and WASH has markedly increased with

additional activities noted, but serious problems with supplies budgets

and material conditions remain.

4. COVID-19 pandemic offers an opportunity to strengthen WASH and IPC

in our hospitals with long term gains beyond the pandemic.

193

Background

Many countries, including Kenya, have noted challenges with Water Sanitation

and Hygiene. A recent global survey in 120 countries by the WHO and UNICEF

revealed that a quarter

of health facilities

assessed lacked water

from an improved

source on the premises

and almost half lacked hand hygiene facilities at the points where care is

provided. With the emergence of COVID-19, proper IPC and WASH have

become essential strategies to minimise the spread of the disease among

patients and health care workers. Recent research has shown that health

workers are at higher risk of contracting COVID-19 in the line of duty due to

inadequate IPC practices. We must adequately protect health workers to avoid

them being harmed and to enable the continuation of services by ensuring IPC

structures are in place in our hospitals.

In 2018, using a modified version of the WHO Water Sanitation and Hygiene

Facility improvement tool,

we conducted surveys of

WASH infrastructure and

organisation in 116 wards

in 14 public hospitals in

Kenya. The Level 4 and 5 hospitals are located in 11 counties in central, eastern

and western regions on Kenya. These surveys were conducted together with

hospital staff members drawn from the 14 hospitals to ensure the hospitals

were equipped to carry out subsequent surveys independently. In addition to

Water Sanitation and Hygiene (WASH): A series

of interdependent components that include the

provision of safe water, proper access to toilets

and waste disposal and hand hygiene facilities

Infection Prevention and Control (IPC): A

scientific approach aimed at preventing the

spread of infections between patients, health

care workers and the community

194

the surveys, in-depth interviews were conducted with hospital managers and

frontline health workers to understand the context better.

After the surveys, individualised written feedback was provided to all the

hospitals enabling them to prioritise necessary improvements. One month after

the first COVID-19 case, we contacted the survey team members to learn what

improvements had taken place and how these hospitals were preparing to deal

with COVID-19 from a WASH/IPC perspective.

Results

Our initial surveys reported marked differences in the performance of WASH

within and across the hospitals. Based on the 65 indicators assessed, spanning

four domains (water, sanitation, hygiene and organisational management) at

the facility level, overall hospital performance varied from 47% to 71% with five

of the 14 hospitals scoring below 60%. There were also significant differences

in performance between wards in the same hospital. Figure 1 illustrates WASH

performance at ward and overall hospital levels.

To improve accountability, we identified three primary levels of accountability

for WASH and IPC in the hospitals. These are IPC committees, hospital managers

and the county government.

195

Figure 1: Hospital and ward performance across the hospitals during the 2018 survey Scatter plot

of the aggregate ward level scores of 116 wards (black shaded points) across 14 hospitals by

domain. The Blue line for each hospital shows average performance in each area across all the

hospitals – this is typically less than 60%. The degree to which the black dots are spread indicates

how large the difference in performance is between wards in the same hospital – in some

hospitals dots are closer together. Hence, performance is more consistent; in others, they are

spread far apart, showing lots of difference.

Some of the poor performance in the hospitals was attributed to limited

resources and funding for IPC and WASH activities in the hospitals:

196

“The challenge is inadequate funding. Now that you are given some money that

cannot meet your expectations, then that one comes as a challenge because

perhaps you want to buy a certain amount of material. However, because of

limited funding, you buy as per what is available." Health Manager

Proper leadership, even at the ward level, was noted to contribute to the

success and excellent performance

in some of the wards and hospitals

surveyed. This leadership plays a

role in how the clinicians operate.

"I think it is because we have

somebody who is very vocal. And she is very strict. She will tell them you are not

going to handle any baby without having washed your hands." Health Manager.

Figure 2 Differences in waste management procedures between hospitals and wards. A shows

colour-coded waste segregation and sharps disposal bins with labels on the wall. B shows the

mixing of infectious and non-infectious waste.

In some of the hospitals, IPC committees were non-functional. In other

hospitals, staff responsible for IPC issues noted this to be a low-status

committee compared to others like the Medicines and Therapeutic Committees

and complained that members did not attend meetings.

A B

“IPC is very important because I think

to me it’s the heartbeat of the

hospital.” Health worker

197

There are some committees which are found to be more,

which are more do I say prestigious? They look better. So, if

I am in IPC, people will be thinking okay… so IPC will have

no one. I mean, what is the benefit of being in IPC, what is

there, how am I gaining being in IPC? Consultant

Other reasons that explained the variation noted in the 2018 survey included

differences in the built environment with some hospitals having a mix of old

and newer wards with better infrastructure. Support for WASH by development

partners in some of the hospitals may also have contributed to better

performance in some of the hospitals.

IPC /WASH in the COVID-19 Pandemic Since the emergence of COVD-19, there had been a sharp focus on IPC activities

from the national government with the issuance of new IPC guidelines to handle

COVID-19. Accordingly, the roles played by the IPC committees have changed

and have gained more importance. Hence, infection control has become an

urgent priority, and the interviews revealed hospital staff at all levels are now

concerned about IPC best practice

In some facilities, this has offered an opportunity to increase the IPC supplies

and infrastructure, including hand hygiene facilities in the hospital.

"I believe that after this we will go far, we never had sanitisers before, these

days we have sanitisers" Hospital IPC leader

Reduced supply of some of

the personal protective

equipment has, however,

increased fear and anxiety

Everybody is worried, but if they provide

people with enough PPE'S, nobody is there to

worry. People are ready to work" IPC Leader.

198

among some of the health care workers.

While there were areas of improvement, the overall assessment suggested that

some of the long-standing issues had not been adequately resolved. These

include challenges with waste management, e.g. broken-down incinerators,

hand hygiene with broken down taps, showers and sinks. Funding for IPC

activities in most hospitals is still inadequate.

Yes, the sinks are prone to blockages, so those were some of the issues we have

been making ways, but because we are given limited cash, they go for the

cheapest, and we have constant blockages. IPC Committee Leader.

Figure 3: Broken Hand hygiene station and hand hygiene posters in the 2018 surveys.

Conclusion

There is a need to prioritise WASH and IPC in public hospitals, especially in light

of the COVID-19 pandemic. Investing in WASH and IPC provides long term gains

beyond the pandemic. It contributes to improving quality of care, patient safety

and patient outcomes.

199

Recommendations

1. We propose frequent WASH surveys to assess, measure and monitor

the progress of WASH in hospitals in resource-limited settings,

providing useful data for decision and policy making and tracking

improvements over time.

2. Increased political engagement from national and county

governments, this provides national standards/guidelines and

accountability mechanisms to ensure that facilities are meeting these

national standards.

3. Health facility managers should also underscore the importance of

Infection Prevention and Control committees, and seek to make IPC

and WASH central to hospital and ward planning.

4. There is an urgent need for adequate resource allocation for WASH and

IPC activities within hospitals to address the COVID-19 pandemic,

including capital funds for the required infrastructure and ongoing

support for adequate PPE and sanitisation materials.

About this Research

Related Publications

This brief is adapted from three research papers published under the titles:

1. "Evaluating the foundations that help avert antimicrobial

resistance: Performance of essential water sanitation and hygiene

functions in hospitals and requirements for action in Kenya".

200

(https://journals.plos.org/plosone/article?id=10.1371/journal.pon

e.0222922),

2. "Extending the use of the World Health Organisations' water

sanitation and hygiene assessment tool for surveys in hospitals –

from WASH-FIT to WASH-FAST

(https://journals.plos.org/plosone/article?id=10.1371/journal.pon

e.0226548).

3. Infection prevention and control during the COVID 19 pandemic;

Challenges and opportunities for Kenyan public hospitals. ( In

Publication).

201

Section 2

Antibiotic use and stewardship in Kenyan

Hospitals

202

203

Chapter 5

Evaluating hospital performance in antibiotic

stewardship to guide action at national and

local levels in a lower-middle-income setting

Jacob McKnight, Michuki Maina, Mathias Zosi, Grace Kimemia,

Truphena Onyango, Constance Schultsz, Mike English and Olga Tosas-

Auguet

Global health action. 2019;12(sup1):1761657.

https://doi.org/10.1080/16549716.2020.1761657

204

ABSTRACT

Background

Inappropriate use of antibiotics can lead to the development of resistant

pathogens. Ensuring proper use of these important drugs in all healthcare

facilities is essential. Unfortunately, however, very little is known about how

antibiotics are used in LMIC clinical settings, nor to what degree antibiotic

stewardship programmes are in place and effective.

Objective

We aimed to record all Antibiotic Stewardship policies and structures in place

in 16 Kenyan hospitals. We also wanted to examine the context of antibiotic-

related practices in these hospitals.

Methods

We generated a set of questions intended to assess the knowledge and

application of antibiotic stewardship policies and practices in Kenya. Using a set

of 17 indicators grouped into four categories, we surveyed 16 public hospitals

across the country. Additionally, we conducted 31 semi-structured interviews

with frontline healthcare workers and hospital managers to explore the context

of and reasons for the results.

Results

Only one hospital had a resourced ABS policy in place. In all other hospitals, our

survey teams commonly identified structures, resources and processes that in

some way demonstrated partial or full control of antibiotic usage. This was

verified by the qualitative interviews that identified common underlying issues.

205

Most positively, we find evidence discipline-specific clinical guidelines have

been well accepted and have conditioned and restricted antibiotic use.

Conclusion

Only one hospital had an official ABS programme, but many facilities had

existing structures and resources that could be used to improve antibiotic use.

Thus, ABS Strategies should be built upon existing practices with national ABS

policies taking maximum advantage of existing structures to manage the supply

and prescription of antimicrobials. We conclude that ABS interventions that

build on established responsibilities, methods and practices would be more

efficient than interventions that presume a need to establish new ABS

apparatus.

KEYWORDS

Antibiotic stewardship; rational drug use; AMR; LMIC; resistance; hospitals

206

BACKGROUND

Antimicrobial Resistance (AMR) represents an existential threat to global

health. The reduction of overuse and improper use of antibiotics through

Antibiotic Stewardship (ABS) programmes is a cornerstone of WHO’s Global

Action Plan (GAP) [1]. The GAP recognises that regulation is weak, that usage

data is poor, and that the situation is made more difficult by a lack of evidence-

based diagnosis and the proliferation of sub-standard antibiotics [1, p. 10]. Poor

prescribing practices combine with low levels of infection prevention and

control to encourage a faster spread of resistant pathogens in Low and Middle-

Income Countries (LMICs) [2]. For example, 214,500 neonatal sepsis deaths

worldwide are thought to be attributable to resistant pathogens, 52% of which

are experienced in just five LMIC countries [3].

More positively, a recent review suggested that ABS interventions in LMICs can

have a positive effect [4]. However, in many cases, studies of interventions

were of low quality with a high degree of heterogeneity in results meaning that

it is difficult to draw direct conclusions from the analysed literature. Two

systematic reviews in higher-income settings also suggested that interventions

can have a positive effect on ABS, but again, the heterogeneity of approaches

makes it difficult to draw conclusions on what works best [5,6].

Few surveys have been conducted in LMIC contexts to show the level of

implementation and effectiveness of existing ABS structures, thinking, and

practices [7]. In addition to providing an important ‘baseline’ for future ABS

interventions, implementation science highlights the importance of

understanding context as a dynamic part of interventions [8]. If ABS is to

become what May calls ‘an ongoing achievement’, we need to understand not

only if an intervention can have an effect, but also why and what the contextual

207

mechanisms are that enable it. Here we report work helping us to understand

context and opportunities for intervention in Kenyan hospitals.

METHODS

Study setting

In Kenya, the Ministries responsible for Health and Agriculture appointed a

multi-sectoral Technical Working Group to develop national guidelines for the

prevention and containment of antimicrobial resistance [9]. Kenya has a

devolved system of government, however, and provision of healthcare falls to

the county governments who are responsible for any implementation of the

AMR policies and plans.

The survey and interviews covered 16 public hospitals in Kenya and were

conducted alongside related work investigating Infection Prevention and

Control [10]. The hospitals are part of the Clinical Information Network (CIN) of

the Kenya Ministry of Health. The CIN was set up to collate data from paediatric

inpatient units to promote development and adoption of evidence-based

clinical guidelines [11] and is coordinated by the Kenya Medical Research

Institute (KEMRI) Wellcome Trust Research Programme. Six of the 16 facilities

are located in areas of Kenya with a high malaria prevalence. The bed capacities

across the 15 county facilities range between 120–550 beds with between 6

and 18 separate inpatient wards and between 5–26 consultants in total across

all medical and surgical specialties [12]. The study hospitals are all public

facilities. Although there are many private hospitals in Kenya (282/724, 39%)

and missionary hospitals (93/724, 13%) public hospitals are thought to manage

a majority of the demand (348/724, 48%) [13]. In one of the facilities, which

acts as the national referral hospital, due to its size, we only assessed ABS

arrangements in the paediatric and neonatal units. These wards were chosen

208

because WASH should be a priority where there are particularly vulnerable

patients and because the research team has long-standing professional

relationships with the management of these wards.

Survey preparation

We sought to identify an ABS survey tool that would sufficiently capture

relevant aspects of ABS in the hospital context. We began with a desk review of

available ABS policies in Kenya and internationally drawing particularly on the

‘U.K.’s NICE guidance as it offered practical guidance at the hospital level. NICE’s

‘Antimicrobial stewardship: systems and processes for effective antimicrobial

medicine use’, was particularly useful in deriving a set of survey questions [14].

It promotes the use of prescription guidelines, a point further emphasised by

NICE’s Infection prevention and control quality standards [15].

The NICE ABS guidance provides a ‘Baseline Assessment Tool’, but this is

developed specifically for the UK NHS. We drew upon our experience in Kenya

to adapt the tool to evaluate extant processes and structures, rather than only

ABS/AMR specific ones. Table 1 (supplementary materials) shows the

references for each question asked. For each of the 17 questions, the surveyor

scores a facility as either fully, partially or not meeting recommended targets

as defined in the study’s standard operating procedures. An open text section

also allows the data collector to record the logic behind the scores assigned.

The ABS data collection tool (Supplementary Materials) was piloted in a hospital

similar to the study hospitals and revised before actual data collection.

Survey data collection

Data collection was conducted by 3 study teams each of 4–5 people in three

different regions of the country: central Kenya (4 hospitals), Western Kenya (5

hospitals), and around the capital city (5 hospitals). The 3 study teams’

members received training from the investigators prior to the data collection

209

process. For each hospital, upon receiving permission from the manager, the

teams approached the head pharmacist. Across the 16 facilities, these were

graduate pharmacists, some of whom had specialised post-graduate training in

clinical pharmacy. The pharmacist was taken through the survey providing

responses that were graded by the study data collection team into one of the

three possible levels.

Survey data analysis

Individual indicators were grouped within the five pre-specified ABS modules:

leadership; accountability and expert support; supplies; monitoring and

reporting; and policy and practice. For each of the 17 indicators ‘heat map’

colour codes were generated based on the indicator: does not meet (red);

partially meets (orange); meets (green). In addition, we compute aggregate

indicator scores by assigning numeric scores to each level. Does not meet the

target, partially meets and meets target were assigned numeric scores 0, 1, and

2 respectively. Percentage scores were derived as a proportion of the sum of

the numerators(individual indicator scores for each hospital) and dividing these

by 32 which is the sum of denominators representing the maximum possible

score if all hospitals met target for each of the indicators.

Qualitative approach

Semi-structured interviews were conducted with hospital managers (e.g.

medical directors, nursing and laboratory heads) and frontline health workers

(e.g. consultants, medical and nursing officers) during the survey visits, in 7 of

the 16 hospitals sampled to ensure spread across different geographical

locations represented by the study hospitals. The survey team included a team

member from each hospital visited, and these individuals made the

introductions and explained the research for each interview.

210

Interviews were conducted by the first, second, fourth and fifth authors in the

hospital setting, using a semi-structured interview instrument that consisted of

different, but strongly related questions to that of the survey [10,12]. All

authors have experience with medical research in Kenya, but the first author is

experienced in using qualitative methods and trained the others in the use of

the semi-structured tool and general interviewing techniques following the

‘long’ or ‘ethnographic’ technique [16]. We conducted 31 interviews between

November 2017 and March 2018 each lasting between 30 mins and 90 mins

with no one refusing to be interviewed or dropping out and no repeats. We

used both purposive and snowball sampling and were mindful of emerging

important issues such as training and managerial support, and experience and

aimed to reflect this diversity in our interviewee sampling strategy.

Responsibilities for drug use are rarely defined by formal rules at the hospital

level, but normative expectations around the roles of different staff are shared

across hospitals in Kenya and well understood by the research team. As such,

we targeted nurses and doctors but also spoke to pharmacists and laboratory

technicians.

The semi-structured interview process provided opportunities to investigate

emergent areas of interest, identified through discussions with the research

team and review of transcripts, and this allowed us to move beyond areas

where we had reached saturation and onto new areas. Interviews were

conducted in a quiet place near the place of work with one or two interviewers

and the interviewee. At the end of each interview, the respondent was asked

to raise any issues that they felt were important but not addressed. We did not

return transcripts to the respondents but did check our findings with some

hospital managers and the Ministry of Health and county stakeholders with

whom we are connected.

211

The audio files were transcribed and uploaded into NVivo 12. Our

methodological framing was exploratory and inductive, and though guided by

initial discussions between the team and the stakeholders, adopted a grounded

theory approach. Open codes were individually generated according to best

practice in qualitative research by the 1st, 2nd and 4th authors [17]. Then we

sought to match the findings of the survey with the themes that emerged from

the qualitative analysis. The analysis presented below is organised so as to allow

us to explore inductively derived issues within the broad categories used in the

survey. This approach allows us to add context and detail to each of the survey

questions.

All interview tools and information sheets are available on request. A

Consolidated Criteria for Reporting Qualitative research (COREQ) checklist was

completed and is included as an appendix, following a format provided by Tong

et al. [18].

RESULTS

The survey was carried out at all 16 hospitals. The most senior pharmacist

available during the visit was chosen to respond to the questions. For the

qualitative interviewing, 31 respondents were interviewed using the semi-

structured interview instrument that was adapted infection prevention

guidelines by ward throughout the research to reflect our growing

understanding of the key issues. Overall the indicator scores ranged between

28 and 69%. The indicator assessing the availability of local speciality-specific

antibiotic guidelines had the lowest overall performance. There was no

significant difference in indicator performance between the hospitals in the

high and low malaria-endemic zones.

212

Leadership

Seven indicators assess leadership for ABS. As can be seen in Figure 1, there was

mixed performance across the hospitals and the indicators, with H5 and H12

being low performers and H14 and H15 being highly compliant. Figure 1 also

indicates that it is unusual for a hospital to have an annual planned budget for

ABS.

Accountability and expert support for ABS

Figure 1 shows that while a person might be identified as being responsible for

ABS, this was far from universal. Additionally, while staff were given some time

to contribute to ABS activities, this was often not an official directive, but

instead a general part of individual workers’ roles. This is explored more

thoroughly in the qualitative section below.

Antibiotic Supply in Hospitals – this simple indicator shows that most hospitals,

experienced difficulties in maintaining a full stock of antibiotics. Encouragingly,

in 4 hospitals, antibiotic supply was not thought to be a problem.

213

Figure 1. Performance at hospital and indicator level across 16 hospitals. The tiles in the central grid are coloured according to the performance classification of each indicator in each hospital, as shown in the colour legend.

214

Monitoring and reporting for ABS activities

We aimed to understand if the pharmacists are able to track how specific

antibiotics are being prescribed across hospital departments or disease

conditions and also monitor the quality of prescriptions, e.g. posology,

duration, and route of administration. Only 1 facility achieved the set targets

for the two indicators assessed. The other facilities were deemed to partially

meet the target because the pharmacists regularly attended the ward rounds

within these departments and kept track of what and how antibiotics are being

prescribed using this informal mechanism.

Policy and practice of ABS

We assessed this through a set of five indicators. As can be seen in Figure 1, we

note significant challenges with guideline availability and use as evidenced by a

lack of speciality-specific guidelines.

Qualitative analysis

ABS leadership

Most of the respondents we spoke to had not heard the phrase Antibiotic

Stewardship. Thus, while they might understand the mechanism, they thought

of ABS as a principle rather than a particular programme or area of work:

“So far no, I think mostly, the majority of the people are yet to

be aware about ‘antimicrobial stewardship programme’. The

majority, even if you do the survey currently, they are not even

aware what this term means.” Nurse and IPC focal person

This general lack of awareness meant that there were no specific budgets for

ABS activities, and also that there was no specific training for ABS. However,

several hospitals demonstrated general leadership in the use of all drugs and

who had managed to fund Continuing Medical Education (CME), and other

215

forms of training focused on the use of antimicrobials. Pharmacists, while often

not familiar with the phrase ‘Antibiotic Stewardship’, were familiar with the

idea of ‘rational use’ of drugs in general and had sought to train other staff on

this issue.

“Usually, we have CMEs in Hospital, and it’s [‘rational use of

drugs’] an issue that is usually shared among hospital staffs.”

Pharmacist

While the interviews and survey identified a number of structures and

resources that could be used for further work in ABS, it is important to note

that leadership must attend to all problems, not just ABS. For example, where

clinical pharmacists do exist, they are well positioned to lead training on this

issue have other issues to deal with, such as intern clinical officers unsure of

general drug use:

“correcting outpatients prescriptions, that actually is the

largest challenge … because for the staff that’s there they

have a very high staff turnover, and patient, and especially

they have a new clinical officers who are on training, and they

rotate therefore I think a span of the three months, and they

are fresh, so there is a lot of hiccups on dosing, right

indications. That is our biggest challenge.” Pharmacist

Accountability and expert support for ABS

It is important to recognise potentially untapped resources and existing

organisational structures in place that could improve prescription practices. The

presence of clinical pharmacists was key because they understood the general

principle and could address inappropriate use of antibiotics where they

216

accepted this responsibility. Thus, in the more effective hospitals, we found

evidence of active and capable medicines and therapeutics committees.

“There’s also a committee in the hospital actually they have

formed that that group. And they normally actually with the …

they involve a lot of departments. So, in the lab we have one

person who normally attend those meetings where they’re

taught actually on how the drugs are supposed to be issued.”

Senior Nurse

But sometimes, even when a medicines and therapeutics committee is in place

and has good representation from different areas of the hospital,

underfunding, and organisational difficulties may mean that tight control of ABS

is beyond them.

“We don’t have a very healthy committee, like a stewardship

to follow up to see this patient has been on the antibiotics for

how long? Because sometimes you can have a prescription of

a patient who has been on ceftriaxone for more than ten days

to me, I don’t think it is proper” Pharmacist

It is also important to recognise the broader spectrum of prescribers working in

Kenyan hospitals. Some nurses, clinical pharmacists and most commonly,

clinical officers who influence prescribing or directly prescribe in Kenya appear

to be excluded from decision making and training.

“I think there are there are some meeting for the therapeutic

committee and maybe that is where they discuss these things.

And then, there’s another committee they call doctors

meetings. You know not everybody is a doctor in this hospital,

217

there are clinical officers, and they are prescribers, they have

pharmaceutical technologists who dispense medicine, and

they’re not doctors, so the, we feel that there’re a group that

are left without understanding what is going on in terms of

therapeutic committee reports they don’t get those reports.”

Lab Manager

A further governance issue detracting from standardised usage of antibiotics is

failure to restrict access of pharmaceutical salespeople who were obviously

successful at promoting the use of their drugs within certain hospitals:

“Maybe with the influence of the med rep and then the many

antibiotics they were prescribing. So, as a committee, we

thought maybe you could narrow these antibiotics”

Pharmacist

Qualitative analysis of antibiotic supplies

The interviews confirmed that the supply of antibiotics was sometimes broken

and though there were clear differences across the sites surveyed, most

respondents recognised this as a problem.

“Stockouts are very often. Specific antibiotics, like even now I

will tell you if a patient needs meropenem currently, we don’t

even have a single vile yeah … Sometimes it causes delay of

treatment because now if a patient needs maybe meropenem,

they will assume maybe they’ve given this patient maybe

ceftriaxone for a while and now the consultants or the

clinician are thinking maybe this patient needs to be changed

go get meropenem from the pharmacy. At that point, they will

218

not get, because at that point is when we initiate now to the

process of procuring meropenem. So that might delay

treatment for one, or two days, which is not good, yeah

because of the protocol also in procurement issues, yeah.”

Pharmacist

In one case, a respondent hinted that supply problems were related to deals

being done with improper or ineffective suppliers.

“They decided to award most of the tenders to the local

suppliers. So that had been the challenge because actually the

local suppliers, most of them don’t have the capacity to supply

the items and sometimes the quality of the drugs they are

supplying, we are not able to assure the quality. So,

sometimes it’s politics ehh? So yes, it’s beyond us.” Pharmacist

Such problems can lead to breaks and delays in treatment which may

contribute to the development of AMR.

“If you really need that drug, for example, if it is something

that I want to use like I have decided this patient have

meningitis and I want to use Ceftriaxone, now Ceftriaxone

runs out by day 2, so the patient buys, in a hospital the patient

buys …Yes, are you seeing the problems that we have. So, they

must have someone they are sending to go and bring. That

somebody might take two days to bring that drug or they tell

you they don’t have money because that is the other thing,

they just tell they don’t have money. And you can look at

them, and you can see clearly, they don’t have money.”

Paediatrician

219

A further important finding suggests that clinician prescribing behaviour

changes when faced with supply shortages. This has particular implications for

people who are adjudged to be unable to afford certain drugs

“Yes, that one you will do, that one you will definitely have to

do. I mean like now if you tell someone like I want to treat

someone, let me say someone has upper respiratory tract

infection and I want to give the Amoxil and I really want to, I

really … me I believe in the original, where I want them to get

the original Amoxil from GSK and the person has Ksh 200.

What are you going to do? You just tell them to go and buy

whatever, I mean with the money you have go and buy what

are they called? You know those bad generic and you let them

do that because what are you going to do? They don’t have

money, you cannot buy for them, and you are not giving them

a solution. So, you give them a solution according to what they

have.” Paediatrician

Hence, poorer patients seem to be more likely to be prescribed medicines that

are known to be less effective, to be unable to afford full treatment courses and

also to find drugs of limited or nil pharmaceutical utility.

ABS monitoring and reporting

As stated above, most structures and professionals that deal with ABS do so as

part of other work, and not due to a particular ABS programme in place in the

hospital. In many locations, there was a practice of the pharmacist advising

clinicians on drug use. This may involve occasional visits to the ward to provide

advice or seeking out particular doctors to discuss cases.

220

“Sometimes, we send the patient back, or sometimes we call

the clinician.” Clinical pharmacist

“But once in a while the pharmacist tends to come in the hall,

you know, they come to the ward sometimes and they will just

… ‘why are you giving this one, why are you not giving this?’…

Even simple things like dosages, they will always call you back

from the ward, you come to the pharmacy, you are told what

is supposed to be the right dosage, yeah.” Medical officer

In sum, however, the auditing practices were very weak, there was little use of

actual data and most processes in place were ad hoc, and involved pharmacists

reacting to particular cases rather than being systematic about usage.

Qualitative analysis of ABS policy and practice

We identified positive policies and practices in place that were not specifically

for ABS, but generally supported ABS. Firstly, the staff we spoke to understood

the usefulness of formularies and there seemed to be a history of attempts to

implement them. Secondly, the large-scale adoption and application of the

national paediatric guidelines show that it is possible to change and standardise

the use of antibiotics in this context. Our respondents identified formularies as

addressing the issues we raised concerning ABS:

“So, we, we have a ceiling from where you can choose which

antibiotics to prescribe, so we have a formulary.” Medical

Superintendent

However, in some locations, our respondents were not sure if they were

inactive use and suggested that their introduction had stalled:

221

“So yes, if it is formulary there is one that it is being

developed. Did they finish developing it? I am not sure. But

there is one that I had been called to actually develop. I am

not sure how far it went. ” Paediatrician

“I remember … actually from … a few years back because I’ve

been here for a while like almost five years, we started

developing a hospital formulary for a whole week, customised

hospital formulary.” Pharmacist

Additionally, we found evidence that where a formulary was not applied, there

was space for differential prescriptions, based not on the specific condition, but

rather on the patient’s insurance status:

“P: If you go to our pharmacies, you will realise that we have

separated patients, some NHIF clients and others. Essentially,

we should just have a uniform formulary that is working for all

of them.

I: You mean that if I got you right that if patient is under

insurance and others paying cash, you separate the formulary

they receive … ?

P: They are likely to get different products sometimes.”

Medical superintendent

The situation was similar with guidelines. Our respondents recognised the

importance of guidelines. They pointed to the success of the paediatric

guidelines, but most often suggested that other guidelines were not in use

anywhere else in the hospitals we visited.

222

“Yah that one would vary, I think it varies for example, we

have a paediatric protocol that have guidelines on how to use

antibiotics. Like this one should not be used in combination

with this one, use these number of days. Maybe that, maybe

that is what we have but in other wards no there is no

guideline. Maybe it is present in pharmacy, maybe pharmacy

have a copy of some sort, but I am not aware”. Medical officer

“No, it is not there but in paediatric wards those ones I know

they are there, but in the wards, I know there is none in the

surgery ward. I don’t know if they have the guideline, but they

use their phones to look at what to give for how many days.”

Nurse

In the absence of a formulary or guideline, clinicians had a variety of coping

mechanisms. Commonly, they searched the internet for answers, but it was not

clear precisely what resources they were drawing on when they did this:

“They use the internet and google” Nurse

Similarly, respondents also pointed to the importance of their training.

“Yes, there are guidelines in this hospital, there … I haven’t

seen a guideline but back at the medical school and our

training here, you need to have a justification for giving

antibiotics, you don’t give everybody antibiotic” Medical

officer

And of consultants to whom they are currently or previously ‘apprenticed’:

223

“No [we do not follow guidelines] but we have our consultant.

Our consultants who are really aggressive on when and what

antibiotics to be given. When to change from first-line

antibiotics when … how do you decide to go on to second-line

and when to go to on third-line or when to refer or when to do

culture? Yes, we have our consultants who are very keen, yes.”

Clinical officer

“Okay, for me like I can said, I still use the protocols for my

former, for my former institution” Medical officer

DISCUSSION

Recent work has highlighted the need to understand the state of antibiotic

stewardship in LMICs [2,3,19] and has sought to provide standardised ways of

approaching this important topic [7]. While evidence interventions can be

effective in LMIC contexts exists, we lack clear benchmarks on current practice

more broadly. Unfortunately, but perhaps not surprisingly, for all but one of the

hospitals we visited in Kenya, there was no official recognition of ABS nor a

specific programme in place to deal with this issue. Moreover, even senior

pharmacists were not familiar with the phrase, though they were familiar with

the more general idea of ‘rational use’ of drugs. Hence, given that the ABS

elements of the 2017–2022 Kenyan national AMR policy [9] are yet to be

implemented by county governments in regional hospitals, all survey questions

that sought specific ABS structures and practices were likely to be scored

negatively.

Despite this, our approach allowed us to understand the current ABS ‘lay of the

land’ and identify extant structures and resources that are being used or could

be used, to control the use of antibiotics. Our findings serve to remind those

224

interested in ABS, that antibiotics are well-established, crucial tools in all health

systems, and so even without a specific ABS programme, their supply and use

is already a matter of concern to a variety of health systems structures and

professionals. In essence, antibiotics have social lives [20] and in many ways,

can be seen as infrastructure [21]. Understanding and accepting the existing

culture of antibiotic use for different geographies is key to interventions in this

area [22].

Many of these issues we uncovered are linked not only to the specific supply

and use of antimicrobials but also to a wide range of health system failings. A

lack of effective human resources; a lack of auditable data; problems in

changing clinician behaviours; and shortages of supplies and budgets are all

chronic issues that undermine many aspects of healthcare with ABS being but

one of them. The origins and causes of these core issues are complex and

beyond the remit of any single ABS intervention but important to tackle if ABS

is to be effective.

Our work illuminates how the low ABS performance captured in our survey may

contribute to AMR. Firstly, inadequate supplies can lead to rationing of

medicines and stark choices regarding the patient’s ability to pay [23]. Clinicians

make value judgements based on class and their perception of a patient’s ability

to pay for drugs outside of the hospital. Secondly, where patients are adjudged

to be poor by a clinician, they may prescribe a cheaper, though possibly less

effective drug or a shorter course (A known behaviour for pharmacists [24]).

Thirdly, in sending patients out of the hospital to find medicines in the private

sector, the clinician understands that this may lead to delays in treatment and

doubts as to whether the drugs the patient or their carer procures are in fact

real and effective rather than fakes or cheap copies. Fourthly, we note a

225

converse effect where a clinician may be inclined to prescribe more expensive

medicines where a patient makes it known that they have national or private

insurance.

These issues are sometimes compounded by other long-standing health system

failings. For example, the regular supply of medicines of dubious quality leads

to a lack of trust in antibiotics, while the minimal availability of effective

microbiological diagnostics, even in large hospitals [25,26], means that doctors

cannot be sure what pathogen they are treating. This could result in a situation

where a clinician may decide that the reason a course of antibiotics has failed

to have the desired effect is due to drug quality rather than AMR.

This strong link to the private sector and doubts of the quality of medicines

reminds us that hospitals are not biomedical vacuums where we can assume

the actions of clinicians to be uniform across different geographies, but rather

organisations entwined in local social and cultural milieus [27] and that the use

of antibiotics in this setting is always connected to, rather than isolated from,

the complex, broad social system of antibiotic use and ‘misuse’ [28].

The survey and related interviews also illuminate opportunities for improved

ABS in Kenya’s hospitals. It was apparent that in paediatrics, the widely

disseminated national guidelines [29] were the dominant influence suggesting

discipline-specific guidelines can influence antibiotic use. Additionally, though

formulary processes were moribund in the hospitals we visited, this was a well-

recognised and potentially important way of guiding practice that could be

reinstated.

We also noted for the hospitals in which they were active, medicines and

therapeutics committees were powerful and influenced not only local

226

prescribing patterns but were also involved in procurement and ensuring the

regular supply of drugs to the facility. This is a particularly important point in

LMICs; more people die due to a lack of medication than to resistant pathogens

[3]. While recent work has sought to set targets for reducing the use of

antimicrobials, most LMICs are well below the new targets [2] suggesting we

should help the patients who need antibiotics to access them in addition to

reducing usage in those who do not.

Further, the study suggested that while few in number, clinical pharmacists are

respected for their ability to provide suitable advice to frontline prescribers on

the appropriate use of antibiotics and could play an important role in ABS

programmes [30].

Limitations

Our survey tool was derived from available U.K. documents but was not

formally validated and is not as comprehensive as the checklist recently derived

by Pulcini et al. [7] although the 4 section, 17 item checklist we used

significantly overlaps with these 7 section 29 item lists. Secondly, the selection

of the hospitals was defined by participation in a clinical information network,

and they may not be representative of all Kenyan hospitals. However, we feel

the combination of a survey tool and multiple interviews have enabled us to

provide important insights into the current status of ABS in Kenyan hospitals

that could inform more specific efforts to promote ABS practices that are suited

to this context.

CONCLUSION

Our findings perform three important functions: the survey offers a first

attempt at benchmarking existing ABS practice in Kenyan public hospitals; the

227

qualitative interviews provide detailed insights into the issues that affect

antibiotic prescribing; and our analysis finds that many existing structures could

be reinvigorated to improve ABS in Kenya. While there has rightly been a lot of

attention given to the importance of ABS in reducing AMR, we should recognise

that the use and misuse of antibiotics is determined by long-standing features

of health systems. Recognising these as a prelude to initiating ABS programmes

may be important to their success in limiting AMR in LMIC contexts.

ACKNOWLEDGEMENTS

We want to thank the Kenyan Ministry of Health and the council of governors

who permitted this work to be carried out. We also thank the hospital

management and clinical teams who supported the work in the survey

hospitals. This work is published with the permission of the Director of the

Kenyan Medical Research Institute (KEMRI).

AUTHOR CONTRIBUTIONS

Jacob McKnight PhD: conception; design; acquisition of data (interviewing);

analysis and interpretation; drafting; revisions.

Michuki Maina MD: conception; design; acquisition of data (interviewing,

survey); analysis and interpretation; drafting; revisions.

Mathias Zosi MSc. : acquisition of data (survey); analysis and interpretation.

Grace Kimemia MSc.: acquisition of data (interviewing); analysis and

interpretation; drafting.

Truphena Onyango MSc.: acquisition of data (interviewing).

Paul Mwaniki MSc.: analysis and interpretation; drafting;

Constance Schultsz PhD: conception; design; analysis and interpretation;

drafting; revisions.

228

Mike English PhD: conception; design; analysis and interpretation; drafting;

revisions.

Olga Tosas-Auguet PhD: conception; design; acquisition of data (survey);

analysis and interpretation; drafting; revisions.

All authors are employed as full-time researchers.

DISCLOSURE STATEMENT

We are aware of no conflicts of interest. The authors alone are responsible for

the views expressed in this publication, and they do not necessarily represent

the views, decisions or policies of their institutions.

ETHICS AND CONSENT

Ethics approval for the work was given by the Kenyan Medical Research

Institute’s (KEMRI) Scientific and Ethical Review Unit. Informed written consent

was secured for all interviewees, and considerable care was taken to ensure the

research did not interrupt work. All information received was handled

confidentially and stored on secure servers. All quotes from the study

respondents were anonymised. This study received approval from the Oxford

Tropical research ethics committee (OXTREC) from the University of Oxford

(Ref: 525–17) and from the Kenyan Medical Research Institute (Ref:

KEMRI/SERU/CGMR-C//086/3450).

FUNDING INFORMATION

MM, G.K., J.M., M.Z. and O.T. were supported by funds through a grant from

the Economic and Social Research Council ESRCS # ES/P004938/1 awarded to

ME A Senior Research Fellowship awarded to M.E. by The Wellcome Trust

(#207522) supported P.M. M Maina received additional support from a grant

229

to the Initiative to Develop African Research Leaders (IDeAL) through the

DELTAS Africa Initiative [DEL-15-003], an independent funding scheme of the

African Academy of Sciences (AAS)’s Alliance for Accelerating Excellence in

Science in Africa (AESA) and supported by the New Partnership for Africa’s

Development Planning and Coordinating Agency (NEPAD Agency) with funding

from the Wellcome Trust [107769/ Z/10/Z] and the U.K. government. The

funders had no role in study design, data collection and analysis, decision to

publish, or preparation of the manuscript

230

REFERENCES

[1] WHO. Global action plan on antimicrobial resistance 2015. Geneva: World Health

Organization; 2017.

[2] Laxminarayan R, Sridhar D, Blaser M, et al. Achieving global targets for antimicrobial

resistance. Science. 2016;353(6302):874–875.

[3] Laxminarayan R, Matsoso P, Pant S, et al. Access to effective antimicrobials: a

worldwide challenge. Lancet. 2016;387(10014):168–175.

[4] Van Dijck C, Vlieghe E, Cox JA. Antibiotic stewardship interventions in hospitals in low-

and middle-income countries: a systematic review. Bull World Health Organ. 2018;96(4):266.

[5] Davey P, Brown E, Charani E, et al. Interventions to improve antibiotic prescribing

practices for hospital inpatients. Cochrane Database Syst Rev. 2013;(4).

[6] Schuts EC, Hulscher ME, Mouton JW, et al. Current evidence on hospital antimicrobial

stewardship objectives: a systematic review and meta-analysis. Lancet Infect Dis.

2016;16(7):847–856.

[7] Pulcini C, Binda F, Lamkang AS, et al. Developing core elements and checklist items for

global hospital antimicrobial stewardship programmes: a consensus approach. Clin Microbiol

Infect. 2019;25(1):20–25.

[8] May CR, Johnson M, Finch T. Implementation, context and complexity. Implement Sci.

2016;11 (1):141.

[9] Kenya Go. National Action Plan on prevention and containment of antimicrobial

resistance 2017-2022. Nairobi, Kenya: Government of Kenya; 2017.

[10] Maina M, Tosas-Auguet O, McKnight J, et al. Extending the use of the world health

organisations’ water sanitation and hygiene assessment tool for surveys in hospitals – from

WASH-FIT to WASH-FAST. PLOS One. 2019;14 (12).

[11] English M, Ayieko P, Nyamai R, et al. What do we think we are doing? How might a

clinical information network be promoting implementation of recommended paediatric care

practices in Kenyan hospitals? Health Res Policy Syst. 2017;15(1):4.

[12] Maina M, Tosas-Auguet O, McKnight J, et al. Evaluating the foundations that help avert

antimicrobial resistance: performance of essential water sanitation and hygiene functions in

hospitals and requirements for action in Kenya. PLOS One. 2019; 14(10).

[13] Kenya, MoH. Nairobi, Kenya: Kenyan Ministry of Health; 2020 cited 2020 Mar 20.

Available from: http://kmhfl. health.go.ke/#/home

[14] NICE. Antimicrobial stewardship: systems and processes for effective antimicrobial

medicine use. London: National Institute for Health and Care Excellence; 2015.

[15] NICE. Infection prevention and control. London: National Institute for Health and Care

Excellence; 2014.

[16] McCracken G. The long interview. Sage; 1988.

[17] Gioia DA, Corley KG, Hamilton AL. Seeking qualitative rigor in inductive research: Notes

on the Gioia methodology. Organizational Res Methods. 2013;16 (1):15–31.

231

[18] Tong A, Sainsbury P, Craig J. Consolidated criteria for reporting qualitative research

(COREQ): a 32-item checklist for interviews and focus groups. Int J Qual Health Care.

2007;19(6):349–357.

[19] Cox J, Vlieghe E, Mendelson M, et al. Antibiotic stewardship in low-and middle-income

countries: the same but different? Clin Microbiol Infect. 2017;23 (11):812–818.

[20] Whyte SR, Van der Geest S, Hardon A. Social lives of medicines. Cambridge University

Press; 2002.

[21] Chandler CI. Current accounts of antimicrobial resistance: stabilisation,

individualisation and anti-biotics as infrastructure. Palgrave Commun. 2019;5 (1):53.

[22] Chandler C, Hutchinson E, Hutchison C Addressing antimicrobial resistance through

social theory: An anthropologically oriented report; 2016.

[23] Miljeteig I, Defaye FB, Wakim P, et al. Financial risk protection at the bedside: how

Ethiopian physicians try to minimize out-of-pocket health expenditures. PloS One.

2019;14(2):e0212129.

[24] Barker AK, Brown K, Ahsan M, et al. What drives inappropriate antibiotic dispensing?

A mixed-methods study of pharmacy employee perspectives in Haryana, India. BMJ Open.

2017;7(3): e013190.

[25] Seale AC, Hutchison C, Fernandes S, et al. Supporting surveillance capacity for

antimicrobial resistance: laboratory capacity strengthening for drug-resistant infections in low-

and middle-income countries. Wellcome Open Res. 2017;2:91. PubMed PMID: 29181453.

[26] Wilson ML, Fleming KA, Kuti MA, et al. Access to pathology and laboratory medicine

services: a crucial gap. Lancet. 2018;391(10133):1927–1938.

[27] Van der Geest S, Finkler K. Hospital ethnography: introduction. Soc Sci Med.

2004;59(10):1995–2001.

[28] Pearson M, Doble A, Glogowski R, et al. Antibiotic pre- scribing and resistance: views

from LMIC prescribing and dispensing professionals. Report to World Health Organisation AMR

Secretariat. 2018.

[29] Kenya, MoH. Basic paediatric protocols for ages up to 5 years. In: Health Mo, editor.

4th ed. Nairobi: Kenyan Ministry of Health; 2016. http://guidelines.health.go.ke/#/category/

27/266/meta

[30] Walia K, Ohri V, Mathai D. Antimicrobial stewardship programme (AMSP) practices in

India. Indian J Med Res. 2015;142(2):130.

232

233

Chapter 6

Antibiotic use in Kenyan public hospitals:

Prevalence, appropriateness and link to

guideline availability

Michuki Maina, Paul Mwaniki, Edwin Odira, Nduku Kiko, Jacob

McKnight, Constance Schultsz, Mike English and Olga Tosas- Auguet

International Journal of infectious diseases 2020. 99: p. 10-18.

https://doi.org/10.1016/j.ijid.2020.07.084

234

ABSTRACT

Objective

To examine prescription patterns and explore to what extent guidelines are

available and how they might influence treatment appropriateness among

hospitalised patients in Kenyan hospitals.

Methods

Data on antimicrobial usage were collected from hospitalised patients across

14 Kenyan public hospitals. For each prescription, appropriateness of treatment

was defined using available local and international treatment guidelines and

through consensus with local medical specialists. Association between

appropriate treatment, guideline availability and other possible explanatory

factors was explored using univariate and multiple regression analysis.

Results

There were 1675 (46.7%) of the 3590 hospitalised patients on antimicrobials

with 3145(94%) of the 3363 antimicrobial prescriptions being antibiotics. Two

patients (0.1%), had treatment based on available antibiotic susceptibility tests.

Appropriate treatment was assessed in 1502 patients who had a single

diagnosis. Of these, 805 (53.6%) received appropriate treatment. Physical

availability of treatment guidelines increased the odds of receiving appropriate

treatment Odds Ratio 6.44[95% CI 4.81-8.64].

Conclusion

Appropriate antibiotic prescription remains a challenge in Kenyan public

hospitals. This may be improved by the availability of context-specific, up-to-

date, and readily accessible treatment guidelines across all the departments,

and by providing better diagnostic support.

235

Keywords

Antimicrobial Resistance, Antibiotics, Point Prevalence Survey

236

BACKGROUND

Antimicrobial resistance (AMR) is an emerging global challenge that is thought to

account for more than 700,000 deaths annually [1]. Certain practices that are

common in resource-limited settings can fuel resistance. Inappropriate use of

antibiotics is highly problematic, especially where they are availed over the

counter without prescriptions. Similarly, the availability of falsified or substandard

drugs is a significant issue [1, 2]. In hospital settings, inappropriate use has been

fuelled by lack of surveillance and diagnostic capabilities, poor antibiotic

stewardship activities and lack of treatment guidelines [3-6].

Efforts to improve the appropriateness of antibiotic use rely on the availability of

relevant usage data. Unfortunately, there is a lack of data on antimicrobial use in

African countries compared to other regions [7, 8]. To measure antibiotic use and

quality of prescriptions at hospital and patient level, point prevalence surveys (PPS)

have been extensively used [9, 10]. They are easy to administer and, in addition to

generating data on antibiotic use, can highlight other problematic aspects of

quality of care, including the quality of prescription. They also offer a useful tool

for audit and feedback on antibiotic use in hospitals to improve decision making

and strengthen antibiotic stewardship activities.

The Government of Kenya in 2017 launched the national plan on prevention and

containment of antimicrobial resistance. One of the strategies is to optimise the

use of antimicrobials. This can be achieved through increasing compliance to

reporting of antimicrobial consumption across the country. [11]

Some studies in Kenya indicate a high prevalence of antibiotic use in both inpatient

and outpatient settings of more than 80%. Also highlighted in these studies are

challenges with the quality of prescriptions, including the use of proprietary drug

names and incomplete prescriptions [10, 12]. These studies were, however, from

237

single hospitals. We set out to conduct an antibiotic point prevalence survey across

14 public hospitals in Kenya, examine prescription quality and explore to what

extent guidelines are available and how they might influence treatment

appropriateness.

METHODS

This study took as its starting point the methods developed for the Global Point

Prevalence Survey of Antimicrobial Consumption and Resistance (GLOBAL-PPS),

which has been widely used and for which tools are freely available. [8] [13].

Setting

The point prevalence survey was carried out between February and April 2018

across 14 public hospitals in Kenya with a total bed capacity of 4152 [14]. These

hospitals are distributed across the central, eastern and western parts of the

country in high and low malaria prevalence regions. HIV prevalence ranges

between 2 and 16% across the counties where these hospitals are located. The bed

capacity, number of specialists, catchment population and the HIV prevalence, are

provided as a supplement. (Supplement 1). These hospitals provide multi-

speciality care and were selected purposively as part of a collaboration between

the Ministry of Health and the KEMRI Wellcome Trust Research Programme [15].

During the surveys, the months of March to May 2018, which are the long rains

period in Kenya experienced heavy rainfall and flooding in most parts of the

country [16]. This period sees increased hospitalisations due to malaria and

diarrhoeal diseases in Kenya [17, 18].

238

Data Collection

Ward-level

At ward level, data were collected on the department type, bed and patient

numbers. These data were used to compute bed occupancy. As a modification to

the global PPS, wards were grouped into five departments, namely adult medical,

adult surgical, paediatric medical, paediatric surgical or neonatal. In facilities with

more than one ward in these groupings, all were surveyed, and results pooled at

group level. The data collection tool is provided as a supplement (Supplement 2).

Patient-level

At the patient level, data were collected on the patients’ age, sex, date of

admission and diagnoses. For the diagnosis, a total of 46 possible options were

provided, 45 of these were categorised by the anatomical system involved. The last

category was when the prescription was not supported by documentation of any

diagnosis for which antibiotics are indicated. This category was labelled as

“Conditions for which antibiotics are not indicated”. Data were also collected on

the antimicrobial type, dose and duration of treatment. Microbiology, antibiotic

susceptibility and biomarker (C-reactive protein; procalcitonin) test results used to

inform the diagnosis and treatment, were also documented. Supplement 3

provides the patient-level data collection tool and the list of diagnoses.

Data Collection Process

Data were collected by clinicians trained on the study methodology. There were

three data collection teams, each with 5-6 members under the leadership of a

clinical pharmacist. The survey followed standard operating procedures and data

capture forms currently in use as part of the Global PPS with some context-

appropriate modifications. One modification included data collection on all

239

weekdays mainly due to logistical reasons. Data were not collected on weekends

and holidays. The survey procedures are given in Supplement 4.

Inclusion and Exclusion Criteria

Ward-level

All wards from paediatric (medical and surgical), adult (medical and surgical

including obstetrics and gynaecology) and neonatal were included in the study.

Psychiatric, Ear Nose and Throat (ENT), renal, neurosurgery, eye and intensive care

units, were not present in all hospitals and were excluded. One large hospital

included in the survey only provides maternal and neonatal care. Data from this

facility were included in relevant analyses.

Patient-level

Data were collected from all hospitalised patients on antimicrobial treatment or

prophylaxis, even if the drug was not administered on the day of the survey.

Outpatients and day-case admissions were excluded.

Electronic Data Entry

At each ward, de-identified data were entered from paper medical records into

REDCap using laptop computers. REDCap® (Research Electronic Data Capture) is a

secure, web-based application that supports data capture [19]. In-built range, error

and validity checks were employed at the point of data entry.

Data Analysis

At ward level, proportions were computed to establish bed occupancy and the

percentage of hospitalised patients on antimicrobial treatment. At the patient

level, since each patient was eligible for a maximum of five diagnoses, specific

diagnoses were computed as a proportion of the total number of diagnoses.

240

For analysis of antimicrobial agents, these were grouped into ten main groups

based on their Anatomical Therapeutic Chemical Classification System (ATC) [20].

In addition to the groups based on the ATC, we generated a new group that was a

combination of benzylpenicillin and gentamicin to capture all the prescriptions in

which this combination was used. This combination is the expected national first-

line treatment for many severe paediatric and neonatal infections in Kenya [21,

22].

Antibiotic treatment appropriateness

In this study, treatment appropriateness is defined as any prescription which is in

keeping with (i) treatment guidelines, (ii) consensus of local expert opinion or (iii)

bacterial speciation and antibiotic susceptibility tests.

To determine treatment appropriateness, we first searched for all available

Kenyan (hospital-based or national) guidelines for each of the 45 diagnostic

classifications in the Global-PPS. For the conditions without local guidance, we

searched for guidelines from the World Health Organisation, the Infectious

Diseases Society of America, the National Institute for Health and Care Excellence

and guidelines from international medical associations, e.g. The European Society

of Intensive Care Medicine. We chose the international guidelines through

consensus with local medical specialists since these guidelines are commonly used

in Kenya for teaching. We documented which of these local/international

guidelines were physically available in the relevant wards during the survey.

De-identified individual patient data including diagnoses and antibiotic

prescription, were then shared with a speciality consultant relevant to the ward

they were admitted to (either a surgeon, physician, obstetrician or a paediatrician).

None of the consultants worked in the participant hospitals, and all were blinded

241

as to the hospital source of the data. Based on the reference guidelines these

consultants reviewed the individual prescriptions and classified treatment as

inappropriate if it was not in keeping with the guidelines, or if it involved redundant

antibiotic combinations (where antibiotics have a similar spectrum of cover). The

principal investigator (PI), identified a sample of ten medical records from each of

the departments and independently assessed appropriateness using the same

approach. This independent, duplicate sample was used to check for congruency

with the decisions made by the specialists. In cases where the decision by the PI

and that of the specialists were not in agreement and consensus was not

forthcoming, the opinion of an infectious disease specialist was sought.

To establish the appropriateness by disease condition, a subset of the data was

used. The data were for those patients with only a single diagnosis that warranted

antibiotic treatment. These individuals were used to calculate the proportions of

appropriate treatment by disease condition.

Factors associated with treatment appropriateness

In keeping with our primary interest, we explored whether the physical availability

of treatment guidelines was associated with treatment appropriateness and

included the number of comorbidities, gender, and duration of hospital stay

(period between admission and day of survey) as additional explanatory factors.

Duration of hospital stay and comorbidities were included in following the

hypothesis that patients that spend longer in hospital are more likely to have a

more significant number of prescriptions, and likely to have more complex illness

[23]. The complexity of the illness and the number of prescriptions may influence

the appropriateness of the prescription choice. Duration of hospital stay was

centred and scaled to improve model stability and reduce the problems of model

242

convergence [24]. Frequency tables and graphs were used to present the data.

Statistical analysis was conducted in R [25].

RESULTS

Data are presented for 14 public hospitals in Kenya. Ten of the hospitals had five

departments (adult medical, adult surgical, paediatric medical, paediatric surgical

and neonatal unit). One hospital (H16) is a maternity hospital, therefore had no

paediatric and adult male units. Hospitals H9, H10 and H11 had no paediatric

surgical units. There were no adult surgical patients hospitalised in H10 during the

survey. A total of 3590 patients were hospitalised during the survey, with a median

of 230 patients, interquartile range (IQR)[137] across the hospitals. The bed

capacity varied across the hospitals with a median of 297 beds IQR[137]. The

median occupancy was 87% IQR[39]. Table 1 provides more description.

243

Adult Medical Adult Surgical Neonatal

Paediatric Medical Paediatric Surgical Total

Hospital Beds

Occupancy (%)

Patient No

No on Antimicrobial

s (%)

Beds Occupancy

(%)

Patient No

No on Antimicrobials

(%)

Beds Occupancy

(%)

Patient No

No on Antimicrobials

(%)

Beds Occupancy

(%)

Patient No

No on Antimicrobials

(%)

Beds Occupancy

(%)

Patient No

No on Antimicrobial

s (%)

Beds Occupancy

(%)

Patient No

No on Antimicrobials

(%)

H1 96 (53.1)

51 36 (70.6) 69 (108.7)

75 53 (70.7) 8 (212.5)

17 16 (94.1) 16 (131.2)

21 18 (85.7) 15 (73.3)

11 11 (100.0)

204 (85.8)

175 134 (76.6)

H2 233 (69.5)

162 48 (29.6) 147 (81.6)

120 24 (20.0) 109 (38.5)

42 17 (40.5) 70 (52.9)

37 17 (45.9) 24 (62.5)

15 3 (20.0) 583 (64.5)

376 109 (29.0)

H3 118 (50.0)

59 25 (42.4) 102 (89.2)

91 73 (80.2) 14 (207.1)

29 17 (58.6) 48 (37.5)

18 15 (83.3) 18 (61.1)

11 2 (18.2) 300 (69.3)

208 132 (63.5)

H4 72 (70.8)

51 15 (29.4) 120 (57.5)

69 21 (30.4) 58 (43.1)

25 1 (4.0) 27 (63.0)

17 12 (70.6) 24 (20.8)

5 3 (60.0) 301 (55.5)

167 52 (31.1)

H5 89 (101.1)

90 33 (36.7) 93 (78.5)

73 28 (38.4) 76 (67.1)

51 14 (27.5) 23 (121.7)

28 22 (78.6) 13 (115.4)

15 6 (40.0) 294 (87.4)

257 103 (40.1)

H6 184 (129.9)

239 52 (21.8) 95 (102.1)

97 34 (35.1) 45 (106.7)

48 15 (31.2) 50 (132.0)

66 53 (80.3) 39 (38.5)

15 3 (20.0) 413 (112.6)

465 157 (33.8)

H7 69 (88.4)

61 41 (67.2) 34 (111.8)

38 34 (89.5) 7 (228.6)

16 16 (100.0) 29 (72.4)

21 19 (90.5) 6 (100.0)

6 3 (50.0) 145 (97.9)

142 113 (79.6)

H8 99 (105.1)

104 47 (45.2) 39 (133.3)

52 39 (75.0) 31 (145.2)

45 20 (44.4) 33 (63.6)

21 18 (85.7) 26 (34.6)

9 8 (88.9) 228 (101.3)

231 132 (57.1)

H9 181 (67.4)

122 41 (33.6) 143 (86.7)

124 66 (53.2) 78 (62.8)

49 18 (36.7) 56 (58.9)

33 32 (97.0) NA NA NA 458 (71.6)

328 157 (47.9)

H10 81 (135.8)

110 47 (42.7) NA NA

NA 35 (105.7)

37 21 (56.8) 38 (176.3)

67 56 (83.6) NA NA

NA 154 (139.0)

214 124 (57.9)

H11 157 (110.2)

173 73 (42.2) 24 (129.2)

31 31 (100.0) 31 (125.8)

39 24 (61.5) 43 (120.9)

52 33 (63.5) NA NA NA 255 (115.7)

295 161 (54.6)

H13 108 (142.6)

154 51 (33.1) 97 (106.2)

103 35 (34.0) 106 (72.6)

77 11 (14.3) 28 (153.6)

43 30 (69.8) 20 (105.0)

21 4 (19.0) 359 (110.9)

398 131 (32.9)

H14 54 (53.7)

29 24 (82.8) 55 (98.2)

54 49 (90.7) 14 (57.1)

8 7 (87.5) 16 (37.5)

6 6 (100.0) 16 (50.0)

8 5 (62.5) 155 (67.7)

105 91 (86.7)

H16 171 (63.2)

108 5 (4.6) 62 (93.5)

58 41 (70.7) 84 (75.0)

63 33 (52.4) NA NA NA NA NA

NA 317 (72.2)

229 79 (34.5)

Total 1712 (88.4)

1513 538 (35.6)

1080 (91.2)

985 528 (53.3)

696 (78.4)

546 230 (42.1) 477 (90.1)

425 331 (77.9)

201 (57.7)

116 48 (41.4) 4166 (86.2)

3590 1675 (46.7)

Table 1: Hospital bed occupancy and antimicrobial prescription proportions

* NA- Indicates this department was absent

244

Overall Antimicrobial Usage

There were 1675 (47%) of the 3590 hospitalised patients on treatment with at least

one antimicrobial agent. Of all the hospital departments, the adult medical

departments had the highest number of hospitalised patients (1513). Of these, 538

(35.6%) were on at least one antimicrobial. In the adult surgical, neonatal,

paediatric medical and paediatric surgical wards there were 990, 546, 425 and 116

patients respectively with 528 (53.3%), 230 (42.1%), 331 (77.9%) and 48 (41.4%)

patients on antibiotics respectively.

Antimicrobial and Antibiotic Use

There were 3363 antimicrobial prescriptions (median 2 per patient). Of these, 3145

(94%) were antibiotic prescriptions, with 95 anti-parasitic agents (3%), 80

antivirals (2%) and 43 antifungal agents (1%) accounting for the remainder. This

report, therefore, focusses on antibiotic prescriptions as they were the most

common.

The 3145 antibiotic prescriptions were made for 1675 patients. These included a

total of 1078, 816, 305, 450 and 92 antibiotic prescriptions in the adult surgical,

adult medical, neonatal, paediatric medical and paediatric surgical units

respectively. Of these total prescriptions, 404 (13%) were a combination of

benzylpenicillin and gentamicin. Cephalosporins were the most common

prescriptions accounting for 772 (26%) of the remaining 2741 prescriptions. They

were followed by nitroimidazole derivatives, mainly metronidazole, with 585 (20%)

prescriptions. Combinations of benzylpenicillin and gentamicin were predominant

in the neonatal unit with 178 (58%) of the 305 prescriptions. The 3rd generation

cephalosporins in the adult medical units represented 263/816 (32%) of

prescriptions. Two (0.1%) of the 1675 patients in the study (one adult medical and

245

one paediatric medical) had treatment based on antibiotic susceptibility tests.

Figure 1 illustrates the proportion of prescriptions by department and hospital.

Figure 1. Bar charts denoting the proportions of drug prescriptions by drug ATC class across 14

hospitals by departments and overall.

Diagnoses warranting treatment

There were 2059 diagnoses (adult medical = 716, adult surgical = 599, paediatric

medical = 449, paediatric surgical = 54 and neonatal 241) recorded from the 1675

hospitalised patients. For adults in the medical units, the most frequent diagnostic

category was “conditions for which antibiotics are not indicated” accounting for

160/716 (21%) of the diagnoses in that department. In paediatrics, the most

common diagnosis was pneumonia or lower respiratory tract infections (148/449

[33%]). To give a clear picture of the disease patterns in these hospital

departments, Figure 2 presents the top ten diagnoses for each department

documented for the 1675 patients and therefore includes some parasitic (malaria)

246

and viral infections like HIV. All other diagnoses in a department are collapsed into

the category “other conditions needing antimicrobial treatment”. Where relevant

the viral and parasitic infections were excluded from the analysis on antibiotic use.

247

Figure 2. Bar Plot illustrating the top ten diagnoses and their proportions (%) for each department. n = total number of diagnoses in each department.

Central Nervous System, URTI; Upper Respiratory tract infection, OBGYN; Obstetrics and Gynaecology

248

Guideline availability and treatment appropriateness

We identified four major local guidelines. A guide on management of common

illnesses in level 4-6 hospitals[26], national guidelines for the treatment of

sexually transmitted illnesses [27], the basic paediatric protocol [22], and a local

hospital guideline for use in the national referral hospital [28]. Others were

from the British Thoracic Society(1 condition), Infectious Diseases Society of

America (10 conditions), the Surviving Sepsis Campaign(2 conditions) and the

National Institute for Health and Care Excellence (2 conditions). Full list is

provided in Supplement 5.

There were 1502( 90%) of the 1675 hospitalised patients,(Adult medical = 421;

adult surgical = 543; paediatric medical = 261; paediatric surgical = 53 and

neonatal = 224 patients), where we could ascertain a final diagnosis. This

allowed for assessment of the per-patient appropriateness of treatment by

condition. Overall, 805(53.6%) of the 1502 patients had appropriate treatment.

The highest percentage of appropriate treatment was in the neonatal

department at 80% [179/224]). Of the 224 patients in the adult medical unit

whose treatment was inappropriate, 140 (63%) prescriptions were for

conditions not requiring antibiotic treatment. This is shown in figure 3

249

Figure 3. Bar chart showing the proportion (%) of patients receiving appropriate treatment by

department based on 1502 patients with a single diagnosis.

Treatment appropriateness by disease conditions

We further explored the disease conditions that were treated appropriately by

the departments from these 1502 patients where a single final diagnosis was

assigned. In the adult medical department, 24 (26%) of the 94 patients with

pneumonia had inappropriate treatment. In the adult surgical unit, skin and soft

tissue infections formed a large proportion of patients, 93/135 (68%) received

inappropriate treatment. This is shown in Figure 4. Patients with a final

diagnosis of conditions where antibiotic treatment was not indicated are

presented as a proportion of the total number of patients.

250

Figure 4. Proportion of treatment inappropriateness by disease conditions. The number on brackets () is the total number of patients with the disease. Those on treatment for Conditions not requiring antibiotics are presented as a proportion of the total number of patients in the department. PUO- Pyrexia of unknown origin, UTI Urinary tract infection

251

Physical Availability of Guidelines

We assessed which treatment guidelines were physically available during the

survey (Supplement 6). The only physically available guideline was the basic

paediatric protocols. These guidelines cover ten conditions, five in each of the

paediatrics and neonates. The five conditions for which guidelines were

available in the neonatal and paediatric units accounted for 94% (210/224) and

56%(147/261) of the patients with a final diagnosis, respectively.

Factors associated with treatment appropriateness

We assessed the main factors that may be associated with treatment

appropriateness in both univariate models and a multiple regression model.

These were duration of hospital stay since admission, gender, physical

availability of guidelines and number of diagnoses.

We ran these models considering two possible outcomes, I. Overall treatment

appropriateness for all patients on antibiotics and II; appropriateness only for

patients with conditions requiring antibiotics.

Both outcomes resulted in the same general pattern of results. With overall

treatment appropriateness for all patients on antibiotics as the outcome, there

was evidence of a negative association of appropriateness of treatment with

increasing the duration of hospital stay. Quadratic terms for the duration of stay

were not statistically significant; hence a linear relationship was assumed.

Physical availability of treatment guidelines increased the odds of receiving

appropriate treatment. There was an increase in the odds of appropriate

treatment with an increase in the number of diagnoses for the population of

patients receiving antibiotics (Table 2).

252

Table 2: Univariate and multivariable models for factors influencing appropriate treatment for patients with conditions requiring antibiotic treatment and for all patients on antibiotics

Univariate Regression Multiple regression

Variable level

Appropriate treatment for

patients with conditions

requiring antibiotics

Appropriate treatment for all

patients receiving antibiotics

Appropriate treatment for

patients with conditions

requiring antibiotics

Appropriate treatment for all

patients receiving antibiotics

OR 95% CI p value OR 95% CI p value OR 95% CI p value OR 95% CI p value

Duration of hospital

Stay 0.63 0.55 0.71 <0.001 0.71 0.63 0.79 <0.001 0.65 0.57 0.74 <0.001 0.71 0.63 0.81 <0.001

Guidelines available

in Hospital Yes 3.73 2.79 4.99 <0.001 6.34 4.78 8.41 <0.001 3.65 2.7 4.95 <0.001 6.44 4.81 8.64 <0.001

Number of

Diagnoses

1 Ref Ref

2 1.12 0.8 1.57 0.671

1.33 0.99 1.8 0.013

1.08 0.76 1.55 0.334

1.28 0.92 1.79 0.005

>2 1.21 0.68 2.15 1.98 1.12 3.51 1.58 0.84 2.99 2.58 1.37 4.87

Gender Male 0.98 0.77 1.24 0.859 1.09 0.88 1.34 0.429 0.84 0.65 1.09 0.198 0.84 0.67 1.07 0.154

253

DISCUSSION

From this report about half the hospitalised patients were on antimicrobial

treatments, mainly antibiotics. The proportions of patients on antibiotics varied

across hospital departments with almost three-quarters of the admitted

paediatric medical patients on treatment. This prevalence of antibiotic use is

consistent with other studies in Africa, with some studies reporting proportions

higher than 70% among hospitalised patients. [12, 29]. Cephalosporins and

penicillins accounted for a substantial proportion of these prescriptions. A

significant increase in the use of penicillins and cephalosporins, particularly

ceftriaxone over the last decade was identified globally [30]. This has been

attributed to factors including economic growth, increased expenditure on

health and increased access to medicines [30]. The use of cephalosporins has

also been increased by its convenient frequency of administration which may

be advantageous on understaffed units and may have lower drug costs

(purchase, preparation and administration) compared to other antibiotics that

require multiple daily doses[31].

From this survey, less than 1% of the antibiotic prescriptions were supported

by laboratory data. Lack of laboratory support influences prescription patterns

and choice [32]. This results in most of the treatment being broad-spectrum,

also encouraging polypharmacy, and may fuel drug resistance [33]. Some

challenges identified in laboratories have been, the high cost of investigations,

long turnaround time and inaccurate results [34]. Improving laboratory capacity

to conduct tests like cultures and antibiotic susceptibility testing would

translate to increase costs of care for the patient. The use of regional referral

laboratories supported by governments to carry out such tests may mean lower

costs for patients due to economies of scale [35]. These extra laboratory costs

254

need to be viewed in light of the unnecessary costs incurred due to

unwarranted treatment and hospital stay in cases where treatment is not

supported by laboratory data.

Our data suggest an association between the physical availability of treatment

guidelines and appropriate treatment. It is striking that across 14 hospitals,

guidelines were only physically available on paediatric and neonatal units.

Better access to approved guidelines for common illnesses is imperative. We

note that the availability and use of the paediatric guidelines have been

supported by other interventions which include nationwide dissemination of

the guidelines by the Ministry of Health and the training of clinicians in hospitals

and universities on the use of these guidelines through the emergency triage

assessment and treatment programme [36, 37]. Adherence to locally relevant

treatment guidelines may improve patient outcomes reducing mortality, the

length of hospital stay and readmissions [38]. Although we assessed for

physical availability of guidelines, we acknowledge that some of the

international guidelines are also available electronically. Availing local

guidelines electronically may increase usage due to the extensive internet

coverage in Kenya.

In addition to the hospital level factors like laboratory and guideline availability,

clinician level factors may also affect antibiotic prescription. Level of training

and clinical experience is crucial in decision making and guideline adherence

[39]. Most of the hospitals in our study were internship training centres. Hence,

prescriptions were done by different cadres of clinicians with varying levels of

training. Though data on prescribers were not collected in this study, work in

similar settings indicates that more than 85% of prescriptions in the paediatric

units are done by junior clinicians mainly in the pre-registration period of

255

training [39]. Therefore, in addition to providing guidelines, improving

supervision of the junior clinicians is crucial[40]. Other interventions that could

enhance proper antibiotic use include functional antibiotic stewardship

committees that can audit antibiotic use and provide feedback to the clinicians

[41].

Our data collection included Mondays in deviation to the Global-PPS

recommendation, especially on surgical prophylaxis data [13]. This was because

patients in our hospitals were admitted for elective surgeries throughout the

week. Additional analysis showed no differences in the prescriptions patterns

and appropriateness by day of the week. Using the PPS approach has some

limitations. It does not capture seasonal variations in disease patterns,

treatment outcomes and does not establish which antibiotics were available in

the hospitals, which may influence the prescriptions. Additionally, other

investigations including, radiology and haematology, which may have informed

the antibiotic prescriptions are not considered in this PPS approach.

We acknowledge the clinical specialties used to assess appropriateness did not

include clinical microbiologists and pharmacists. While including them would

have been ideal, we consulted an infectious disease specialist to assist where

the decisions on appropriateness were not clear. Additionally, the guidelines

we included provided different treatment approaches for similar illnesses

hence accommodating the differences in antibiotic selection between the

prescribers.

Our assessment of appropriateness did not consider the doses and duration of

treatment. However, we note that in more than 96% of the prescriptions, the

dose and duration were documented. Finally, this study was performed in

256

public hospitals, although these are the majority, it does not inform about

appropriateness of prescription and usage in private hospitals in Kenya.

CONCLUSION

We report that 46.7% of all inpatients were on antimicrobial treatment, mainly

antibiotics. In <0.1% of patients, the antibiotic prescription was supported by

laboratory tests. Many patients did not have a recorded diagnosis that

warranted an antibiotic prescription, and there is a strong indication that

physical availability of guidelines may influence treatment appropriateness

positively. This situation may be improved by availing treatment guidelines

across all the departments and providing better diagnostic support and training

for clinicians.

Authors’ contributions

The roles of the contributors were as follows: M.M, J.M, O.T, C.S and M.E

conceived the study. M.M E.O and N.K reviewed the prescription data. M.M and

P.M conducted data analysis. M.M, J.M, O.T, C.S and M.E drafted and critically

revised the manuscript for intellectual content. All authors read and approved

the final manuscript.

Acknowledgements

The authors would like to thank all the hospitals that participated in the survey.

We also thank all the clinicians involved in data collection. This work is

published with the permission of the director of KEMRI.

Funding

This work was supported by funds from the economic and social research

council ESRCS # ES/P004938/1, and a Senior Research Fellowship awarded to

257

ME by The Wellcome Trust (#207522). MM is supported by a grant from by the

Initiative to Develop African Research Leaders (IDeAL) through the DELTAS

Africa Initiative [DEL-15-003], an independent funding scheme of the African

Academy of Sciences (AAS) ’s Alliance for Accelerating Excellence in Science in

Africa (AESA) and supported by the New Partnership for Africa’s Development

Planning and Coordinating Agency (NEPAD Agency) with funding from the

Wellcome Trust [107769/Z/10/Z] and the UK government. The funders had no

role in drafting nor the decision for submitting this manuscript.

Competing interests

There are no conflicts of interest declared by the authors.

Ethical approval

This study received approval from the Oxford Tropical research ethics

committee (OXTREC) from the University of Oxford (Ref: 525–17) and the

Kenyan Medical Research Institute (Ref: KEMRI/SERU/CGMR-C//086/3450).

Availability of data and materials

Data used for this manuscript are available in Harvard Dataverse at

https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/

L7S8TK. Applications for access can be made through the Data Governance

Committee with details available on www.kemri-wellcome.org, or on email to

cgmrc@kemri-wellcome.org.

258

REFERENCES

1. O’neill, J., Antimicrobial resistance: tackling a crisis for the health and wealth of

nations. Review on antimicrobial resistance, 2014. 1(1): p. 1-16.

2. Ayukekbong, J.A., M. Ntemgwa, and A.N. Atabe, The threat of antimicrobial

resistance in developing countries: causes and control strategies. Antimicrob Resist

Infect Control, 2017. 6: p. 47.

3. Petti, C.A., et al., Laboratory Medicine in Africa: A Barrier to Effective Health Care.

Clinical Infectious Diseases, 2006. 42(3): p. 377-382.

4. Elbireer, A.M., et al., The Good, the Bad, and the Unknown: Quality of Clinical

Laboratories in Kampala, Uganda. PLOS ONE, 2013. 8(5): p. e64661.

5. Doron, S. and L.E. Davidson, Antimicrobial stewardship. Mayo Clin Proc, 2011.

86(11): p. 1113-23.

6. Paterson, D.L., The Role of Antimicrobial Management Programs in Optimizing

Antibiotic Prescribing within Hospitals. Clinical Infectious Diseases, 2006.

42(Supplement_2): p. S90-S95.

7. Ahoyo, T.A., et al., Prevalence of nosocomial infections and anti-infective therapy

in Benin: results of the first nationwide survey in 2012. Antimicrob Resist Infect

Control, 2014. 3: p. 17.

8. Versporten, A., et al., Antimicrobial consumption and resistance in adult hospital

inpatients in 53 countries: results of an internet-based global point prevalence

survey. The Lancet Global Health, 2018. 6(6): p. e619-e629.

9. Gharbi, M., et al., Using a simple point-prevalence survey to define appropriate

antibiotic prescribing in hospitalised children across the UK. BMJ Open, 2016. 6(11):

p. e012675.

10. Okoth, C., et al., Point prevalence survey of antibiotic use and resistance at a

referral hospital in Kenya: findings and implications. Hosp Pract (1995), 2018.

46(3): p. 128-136.

11. Republic of Kenya, National Policy for the Prevention and Containment of

Antimicrobial Resistance. 2017: Nairobi, Kenya.

12. Mulwa, N.C., et al., Patterns of Prescribing Practices in Makueni County Referral

Hospital, Kenya. African Journal of Pharmacology and Therapeutics, 2015. 4(4).

13. GLOBAL-PPS. Global Point Prevalence Survey of Antimicrobial Consumption and

Resistance 2015 [cited 2017 16th February]; Available from: http://www.global-

pps.com/.

14. Maina, M., et al., Evaluating the foundations that help avert antimicrobial

resistance: Performance of essential water sanitation and hygiene functions in

hospitals and requirements for action in Kenya. PLoS One, 2019. 14(10): p.

e0222922.

259

15. English, M., et al., What do we think we are doing? How might a clinical information

network be promoting implementation of recommended paediatric care practices

in Kenyan hospitals? Health Research Policy and Systems, 2017. 15(1): p. 4.

16. Ministry of Environment and Forestry, Review of rainfall during the March to May

2018 “Long rains”season and the outlook for the June-July-August (jja) 2018, State

Department of Environment, Editor. 2018, Kenya Meteriological Department,:

Nairobi.

17. Apat, D.O., et al., Temporal variation in confirmed diagnosis of fever-related

malarial cases among children under-5 years by community health workers and in

health facilities between years 2013 and 2015 in Siaya County, Kenya. Malaria

Journal, 2017. 16(1): p. 454.

18. Tornheim, J.A., et al., The epidemiology of hospitalization with diarrhea in rural

Kenya: the utility of existing health facility data in developing countries.

International Journal of Infectious Diseases, 2010. 14(6): p. e499-e505.

19. Harris, P.A., et al., Research electronic data capture (REDCap)--a metadata-driven

methodology and workflow process for providing translational research

informatics support. J Biomed Inform, 2009. 42(2): p. 377-81.

20. World Health Organization, Guidelines for ATC classification and DDD assigment.

2015.

21. World Health Organization, Pocket Book of Hospital Care for Children: Guidelines

for the Management of Common Childhood Illnesses. 2013, World Health

Organization Geneva.

22. Ministry of Health. Basic Paediatric Protocols for ages up to 5 years. . 2016 [cited

2019 12th July ]; Available from:

http://guidelines.health.go.ke/#/category/27/266/meta.

23. Nobili, A., et al., Polypharmacy, length of hospital stay, and in-hospital mortality

among elderly patients in internal medicine wards. The REPOSI study. European

Journal of Clinical Pharmacology, 2011. 67(5): p. 507-519.

24. Pinheiro, J.C., D.M. Bates, and M.J. Lindstrom, Model building for nonlinear mixed

effects models. 1995: University of Wisconsin, Department of Biostatistics

Madison, WI.

25. R Core Team, R: A language and environment for statistical computing. R

Foundation for Statistical Computing, Vienna, Austria. 2013.

26. Ministry of Medical Services and Ministry of Public Health and Sanitation, Clinical

Management and Referral Guidelines – Volume III: Clinical Guidelines for

Management and Referral of Common Conditions at Levels 4–6: Hospitals. 2009,

Ministry of Medical Services and Ministry of Public Health and Sanitation,: Nairobi.

27. Ministry of Health, Kenya National Guidelines for Prevention, Management and

Control of Sexually Transmitted Infections. 2018, National AIDS and STI Control

Programme: Nairobi.

260

28. Kenyatta National Hospital and University of Nairobi, The KNH guide to empiric

antimicrobial therapy 2018: Nairobi.

29. Chukwuani, C.M., M. Onifade, and K. Sumonu, Survey of drug use practices and

antibiotic prescribing pattern at a general hospital in Nigeria. Pharmacy world &

science : PWS, 2002. 24(5): p. 188-195.

30. Van Boeckel, T.P., et al., Global antibiotic consumption 2000 to 2010: an analysis

of national pharmaceutical sales data. The Lancet Infectious Diseases, 2014. 14(8):

p. 742-750.

31. Davis, R. and H.M. Bryson, Ceftriaxone. PharmacoEconomics, 1994. 6(3): p. 249-

269.

32. Chem, E.D., D.N. Anong, and J.K.T. Akoachere, Prescribing patterns and associated

factors of antibiotic prescription in primary health care facilities of Kumbo East and

Kumbo West Health Districts, North West Cameroon. PLoS One, 2018. 13(3): p.

e0193353.

33. Chokshi, A., et al., Global Contributors to Antibiotic Resistance. J Glob Infect Dis,

2019. 11(1): p. 36-42.

34. Alemnji, G.A., et al., Strengthening national health laboratories in sub-Saharan

Africa: a decade of remarkable progress. Trop Med Int Health, 2014. 19(4): p. 450-

8.

35. Ministry of Health, Kenya Health Sector Referral Strategy (2014–2018), Division of

Emergency Disaster Risk Management, Editor. 2014.

36. Irimu, G., et al., Developing and introducing evidence based clinical practice

guidelines for serious illness in Kenya. Arch Dis Child, 2008. 93(9): p. 799-804.

37. English, M., et al., Developing guidelines in low-income and middleincome

countries: Lessons from Kenya. Archives of Disease in Childhood, 2017. 102: p.

archdischild-2017.

38. Wathne, J.S., et al., The association between adherence to national antibiotic

guidelines and mortality, readmission and length of stay in hospital inpatients:

results from a Norwegian multicentre, observational cohort study. Antimicrobial

Resistance & Infection Control, 2019. 8(1): p. 63.

39. Ogero, M., et al., Examining which clinicians provide admission hospital care in a

high mortality setting and their adherence to guidelines: an observational study in

13 hospitals. Archives of Disease in Childhood, 2020: p. archdischild-2019-317256.

40. Papoutsi, C., et al., Social and professional influences on antimicrobial prescribing

for doctors-in-training: a realist review. Journal of Antimicrobial Chemotherapy,

2017. 72(9): p. 2418-2430.

41. Wei, X., et al., Effect of a training and educational intervention for physicians and

caregivers on antibiotic prescribing for upper respiratory tract infections in children

at primary care facilities in rural China: a cluster-randomised controlled trial. The

Lancet Global Health, 2017. 5(12): p. e1258-e1267.

261

Supplement 1: Hospital capacity and population demographics

a Kenyan Ministry of Health. "Kenya HIV estimates report 2018." (2018). b 2019 Kenya Population and Housing Census: Distribution of

Population by Administrative units http://www.knbs.or.ke

Facility Hospital Bed

capacity

Number of

specialist doctors

(consultants)

Number of

wards in the

facility

County HIV

Prevalence a

Catchment

population b

Wards Excluded by Specialty

High Malaria Prevalence Zone

H1 204 5 7 7.7% 142,408 Renal Unit

H3 550 12 14 4.5 188,212 Psychiatry, Renal Unit, ENT

H7 180 7 6 16.3% 220,997 None

H8 250 14 8 4.3% 147,992 Renal Unit

H14 165 5 5 5.4 95,292 None

Low Malaria prevalence Zone

H2 594 26 16 2.8 127,100 ENT, ICU, Psychiatry, Renal

H4 216 8 7 3.7 99,065 None

H5 231 7 9 3.1 122,740 Psychiatry

H6 383 17 10 4% 145,903 Neurosurgery

H9 550 19 18 3.8 170,606 ENT, Renal, Psychiatry, ICU

H10 131 24 6 6.1 988,808 None

H11 320 21 9 6.1 197,489 None

H13 378 20 15 3.7 140,338 ENT, Renal, Psychiatry, ICU, Eye

H16 350 14 6 6.1% 268,276 None

262

Supplement 2: Ward Level Data Collection Tool

Please fill in one form for each ward and department type identified within the ward

Unique ID ____________________________________

Date of survey (dd/mm/year) ___________/________/________________

Auditor code (Person completing

form) ____________________________________

Hospital name ____________________________________

Ward name □ Adult Medical (1-6) □ Orthopaedics (1-6) □ Gynaecology (1-6) □ Other:_____________________

□ Adult Surgical (1-6) □ Paediatrics (1-6) □ Maternity (1-6)

□ Surgical (1-6)

□ Neonatal (1-6) □ Post-Natal (1-6)

Does the ward have patients from

mixed departments?

□ Yes □ No

Choose department type: □ Adult Medical □ Paediatric Medical □ Neonatal

□ Adult Surgical □ Paediatric Surgical

Choose patient population for this

department type □ Male □ Female □ Paediatric / Neonatal

Total number of admitted patients present on this ward, for the selected department type, and who were admitted before 8 am: _

Total number of beds on this ward for the selected department type, which were present at 8 am on the day of the survey: _

263

NOTES

• Fill in one form for each ward and department type identified within the ward.

• Date of Survey refers to the date for the current data abstraction – click today on the calendar.

• 'Unique ID' is automatically generated by the REDCap when you click this field. This is a unique code used to identify each ward form record.

• If the name of the ward is other than those listed, choose 'other' and enter ward name and number (i.e. either 1, 2, 3 …6) to identify this ward

uniquely

• Include only inpatients "admitted before 08:00 o'clock" on the day of the survey

• For numerical fields, type -1 if the value is unknown

• Orthopaedic and neurosurgery patients correspond to the surgical department type

• DO NOT conduct a survey of the following wards: Amenity, Psychiatry, ENT, Renal, ICU, NICU (including in Kenyatta hospital), Neurosurgery and

Eye

• In the case of paediatrics and neonatal, beds may correspond to cots or incubators. Cots and incubators must be included in the total 'bed

count' where applicable.

264

Supplement 3: Patient-level data collection tool and diagnosis list

Please fill in one form per patient on antibiotic or other eligible antimicrobial treatment/prophylaxis—explanatory Notes I explains how to fill this form.

Unique Patient

IP Hospital

Name Ward Name Activity

Today's Date

Admission Date

Date of Surgery I

Date of Surgery II

Re-admission

ID 1 Number

2 Name/Code Name/Code

(Department Type) 3

--/--/----4

--/--/----5 --/--/---- 6 --/--/---- 7 Y □ / N □ / Empty □ 8

Is Patient’s age documented? 9

Patient Age 10 : ≥ 2 years □ / 1-23 months □/ <1 month □

Weight 11 Gender

12 Diagnosis (Explanatory

Note II) 13

Y □ / N □ Years

(if ≥ 2 years): ____________

Months (1-23 months): _____________

Days (if <1 month):

_____________

(in Kg; 1 decimal)

□ M □ F

□ Diagnosis 1

□ Diagnosis

4

□ Diagnosis 2

□ Diagnosis

5

□ Diagnosis 3

Treatment based on biomarker data? 14

Y □ / N □ If yes, which biomarker? 15 CRP □ / PCT □ / Other □ Number of antimicrobials

given 18

Type of biological fluid sample? 16

Blood □ / Urine □ / Other □

Most recent value of biomarker (mg/l)? 17

____________________ (from 0 to 5)

265

Antibiotic or other antimicrobial name: 19 _____________________________

Start Date: 20

__/__/____ Type of indication (Explanatory

note IV) 25 _____________________________

Single Unit Dose: 21

__________ Reason in notes 26 Y □ / N □

Unit: 22 g □ / mg □ / mega unit □ / tablet □/ IU □/ empty □

Guideline compliance 27 Y □ / N □ / NA □ / NI □

Doses per day: 23

OD □ / BD □ / TID □ / OID □/ empty □/ Other □:__________

Stop/Review date documented: 28

Y □ / N □

Route: 24 P □ / O □ / R □ / I □/ empty □/ Other □:__________

Duration: 29 _____________________________

Treatment (E: Empirical / T: Targeted) 30

E □ / T □

The next section is to be filled in only if the treatment of choice is based on microbiology data (i.e. treatment = targeted):

MRSA 31 Y □ / N □

MRConS 32 Y □ / N □

VRE 33 Y □ / N □

ESBL-producing enterobacteriaceae 34 Y □ / N □

Third generation cephalosporin resistant enterobacteriaceae / non-ESBL or ESBL status unknown 35

Y □ / N □

Carbapenem-resistant enterobacteriaceae 36 Y □ / N □

ESBL-producing non-fermenter Gram negative bacilli 37 Y □ / N □

Carbapenem-resistant non-fermenter Gram negative bacilli 38 Y □ / N □

Targeted treatment against other MDR organism 39 Y □ / N □

266

Explanatory Notes I – Patient Form Adapted from the Global Point Prevalence Survey (2017 GLOBAL-PPS):

http://www.global-pps.com/documents

1 Unique ID: This is a 'survey number', a unique non-identifiable number given

by the electronic system for each patient entered in the database. Note down

this number with the corresponding 'Patient IP Number' every time you enter

patient information in the database. This will allow you to retrieve the record

for this patient again; in the event, you need to amend information or record

additional information. You can only search patients in the database, using the

survey number. The unique ID is a seven-digit auto-generated and incremental

number. For example; Kiambu, the first record would read 5100001.

2 Patient IP Number: A unique inpatient identifier that allows linkage to

patient records at the local level for a more detailed audit. This unique identifier

will eventually be removed from the database after the data has been cleaned,

and before the analysis starts. Check all the medical record notes, including file

covers, for the IP number. 3 Activity: This is the department type. E.g. M=medicine, S=surgery (including

orthopaedics, and any other surgical patients in obstetrics or other), etc.

4 Today's Date: date when records are abstracted. Click on 'today' icon to

generate the date for today.

5 Admission Date: Please indicate the date of admission as indicated on the

ward clinical admission notes- it is the date when the patient is first seen by a

clinician. If not available check in other sections of the file (e.g. cardex) or ask

the ward staff.

6 Date of Surgery I: This question is only applicable for patients in a surgical

department. If a surgical department was indicated under activity, this section

will appear in the electronic system. Please indicate the date when the patient

first underwent surgery during the current admission, as indicated on any of

the clinical surgical notes. If not available check in other sections of the file (e.g.

cardex, theatre notes). If date of surgery is definitely missing, enter 01/01/1913

7 Date of Surgery II: As above, if a surgical department was indicated under

activity, this section will appear. Please indicate the date when the patient

underwent subsequent surgery during the current admission, as indicated on

any of the clinical surgical notes. If not available check in other sections of the

file (e.g. cardex, theatre notes). If the patient underwent subsequent surgery

but the date is definitely missing, enter 01/01/1913. If the patient did not

undergo subsequent surgery, enter 01/01/1914. This means 'not applicable'

267

8 Re-admission: Refers to re-admission to this hospital within the 6 months

preceding today's date. Please indicate yes if the patient has been admitted in

this hospital within the previous 6 months. You can check for this information

in the discharge summary for the previous admission. If not documented, enter

Empty.

9 Is patient's age documented? Enter yes or no. If yes, a drop box prompting

to specify the age group will appear in the system.

10 Patient's Age: Choose whether patient is <1 month; 1-23 months old or ≥ 2

years old. If the patient is 2 years old or older, the system will ask you to enter

age in years; if between 1 and 23 months you will be asked to enter the number

of months (e.g. 1 to 23), if less than 1 month you will be ask to enter the number

of days. To report age in years, round to the lowest number (e.g. 2 years and

one month = 2 years; 2 years and 11 months = 2 years). If the information is not

available, enter -1.

11 Weight: This refers to admission weight. Round off weight to the nearest one

decimal place e.g. 6.750 kg = 6.8 kg. Check from the admission notes. If missing,

check on the treatment sheet or nursing cardex. If two readings for the

admission day are different, enter the one on the treatment sheet. If the weight

is not documented, please indicate -1.

12 Gender: Please check from the admission notes. If not documented check

from the records registration form. Alternatively, you can ask a member of staff

to check on the gender of the patient, as he/she will be physically in the ward.

13 Diagnosis/ses: These are what clinicians aim to treat. See diagnoses groups

list in Explanatory Notes II. You may enter up to five diagnoses by order of

priority (i.e. diagnosis 1 would correspond to the main or most severe

condition). You must then select the diagnosis group (as shown in Explanatory

Notes II), that best describes the diagnosis recorded in the patient's clinical

notes.

14 Treatment based on biomarker data: Tick yes if CRP (C-reactive protein), PCT

(Procalcitonin) or culture results are available in clinical notes for this patient.

Click no if none of these laboratory tests were conducted for this patient.

15 If yes, which biomarker: If treatment based on biomarker, specify which one:

CRP (C-reactive protein), PCT (Procalcitonin) or Other (=lab-based culture and

sensitivity result from a relevant biological sample). You may choose more than

one option.

268

16 Type of biological fluid sample: Indicate which type of sample was used to

measure the biomarker (either blood, urine or other [e.g. cerebrospinal fluid,

pus, etc.]). You may choose more than one option.

17 Most recent value of biomarker (mg/l): If CRP (C-reactive protein) or PCT

(Procalcitonin) results are available, write the value of the latest test result in

mg/l.

18 Number of antimicrobials given: Enter the total number of antimicrobials

given to the patient, counting only those antimicrobial treatments that meet

the eligibility criteria described in the outline protocol. If no antimicrobials were

given to the patient, enter zero.

19 Antibiotic or other antimicrobial name: Select the generic antibiotic or

antimicrobial name from the dropdown list. Please note the dropdown list does

not display brand names. You must find out the generic name in case a brand

name is noted in the clinical files instead.

20 Start date: Indicate the date when the treatment started. If this is not

available, enter 01/01/1913

21 Single Unit Dose: Numeric value for dose per administration (in grams,

milligrams, megaunits, tablets or IU). Please see Explanatory Notes III for

reporting specific drug combinations. If dose is missing enter -1.

22 Unit: The unit for the dose (g, mg, mg, tablet or IU) as documented in the

treatment sheet. Indicate empty if not documented.

23 Doses per day: This is the number of times given per day as per the

dropdown list provided in the data capture system, and as documented by

clinicians in the treatment charts or clinical notes (e.g. OD/once a day/24hrly or

BD/twice a day/12hrly or TID/TDS/three times a day/8hrly or

QID/4quarterly/6hrly or STAT [i.e. once only]).If the doses are not as described

above choose 'other' from the dropdown list, and enter the dose as free text. If

necessary provide fractions of doses: (e.g., every 16h = 1.5 doses per day, every

36h = 0.67 doses per day, every 48h = 0.5 doses per day). If doses are not

documented at all choose empty.

24 Route: Indicate the route of administration as documented by the clinician.

Routes of administration are: Parenteral (P [e.g. iv, im]), Oral (O), Rectal (R),

Inhalation (I). Indicate empty if not documented. If route of administration is

none of the above choose 'other' from the dropdown list and enter route as

free text.

25 Type of Indication: Based on diagnosis and/or clinical signs and symptoms

documented in clinical notes, choose the indicator that best describes the

269

indication for the current antimicrobial treatment., See Indication codes in

Explanatory Notes IV.

26 Reason in Notes: Indicate whether the diagnosis / indication for treatment

is recorded in the patient's documentation (treatment chart, notes, etc.) at the

start of antibiotic or other antimicrobial treatment (Yes or No)

27 Guideline Compliance: Refers to whether the antibiotic or antimicrobial

choice (not route, dose, duration etc) follows local guidelines (Y: Yes; N: No; NA:

Not assessable because no local guidelines for the specific indication; NI: no

information because indication is unknown)

28 Stop/Review date documented: Choose 'Yes' if a stop review date is

documented or if the duration of the treatment is indicated in the treatment

sheet or the clinical notes. If the duration of treatment is indicated, this will be

indicated as number of days or weeks or in the following form: X/7, X/52, etc.

Choose 'No' if neither a stop/review date nor duration of treatment are

documented.

29 Duration: Corresponds to the number of days for which the treatment has

been prescribed, as shown by the number of days ticked on the treatment sheet

(i.e. intended duration), even if not all prescribed doses have been given or if

doses have been missed. If duration is not documented enter -1.

30 Treatment Type: Indicate whether the treatment of choice is supported by

microbiology data (i.e. culture results and drug susceptibility test [DST] results

available). If culture results and DST are available, indicate that the treatment

is Targeted (T), if not indicate the treatment is Empirical (E). If Targeted, a list of

organisms will appear, and you will need to indicate which of the organisms

were identified, by ticking 'Yes' or 'No' for each organism listed:

31 Methicillin-resistant Staphylococcus aureus (MRSA)

32 Methicillin-resistant coagulase negative staphylococci (MRCoNS)

33 Vancomycin-resistant enterococci (VRE)

34 Bacteria, producing extended-spectrum beta-lactamases (ESBL)

35 Third generation cephalosporin resistant enterobacteriaceae / non-ESBL or ESBL status unknown

36 Carbapenem-resistant Enterobacteriaceae (CRE) – enteric bacteria resistant to imipenem, meropenem or other carbapenems

37 Nonfermenters: Pseudomonas aeruginosa, Acinetobacter baumannii, Burkholderia spp., Stenotrophomonas maltophilia

38 Carbapenem-resistant Nonfermenters (CR-NF) – nonfermenters resistant to imipenem, meropenem or other carbapenems

39 Multi-drug resistant (MDR) pathogens, others than the listed above.

270

Explanatory Notes II – Diagnostic Codes

Adapted from the Global Point Prevalence Survey (2017 GLOBAL-PPS):

http://www.global-pps.com/documents

Diagnostic Codes (What the Clinician Aims at Treating)

Site Codes Examples

CNS Proph CNS Prophylaxis for CNS (neurosurgery, meningococcal)

CNS Infections of the Central Nervous System

EYE Proph EYE Prophylaxis for Eye operations

EYE Therapy for Eye infections e.g., Endophthalmitis

ENT Proph ENT Prophylaxis for Ear, Nose, Throat (Surgical or Medical

prophylaxis=SP/MP)

ENT Therapy for Ear, Nose, Throat infections including mouth,

sinuses, larynx

RESP Proph RESP Pulmonary surgery, prophylaxis for Respiratory pathogens e.g.

for aspergillosis

LUNG Lung abscess including aspergilloma

URTI Upper Respiratory Tract viral Infections including influenza but

not ENT

Bron Acute Bronchitis or exacerbations of chronic bronchitis

Pneu Pneumonia or LRTI (lower respiratory tract infections)

TB Pulmonary TB (Tuberculosis)

CVS Proph CVS Cardiac or Vascular Surgery, endocarditis prophylaxis

CVS CardioVascular System infections: endocarditis, endovascular

prosthesis or device e.g.

pacemaker, vascular graft

GI Proph GI Surgery of the Gastro-Intestinal tract, liver or biliary tree, GI

prophylaxis in neutropaenic

patients or hepatic failure

GI GI infections (salmonellosis, Campylobacter, parasitic,

C.difficile, etc.)

IA Intra-Abdominal sepsis including hepatobiliary, intra-

abdominal abscess etc.

SSTBJ Proph BJ Prophylaxis for SST, for plastic or orthopaedic surgery (Bone or

Joint)

SST Skin and Soft Tissue: Cellulitis, wound including surgical site

infection, deep soft tissue

not involving bone e.g., infected pressure or diabetic ulcer,

abscess

271

BJ Bone/Joint Infections: Septic arthritis (including prosthetic

joint), osteomyelitis

UTI Proph UTI Prophylaxis for urological surgery (SP) or recurrent Urinary

Tract Infection (MP)

Cys Lower UTI

Pye Upper UTI including catheter related urinary tract infection,

pyelonephritis

GUOB Proph OBGY Prophylaxis for OBstetric or GYnaecological surgery

OBGY Obstetric/Gynaecological infections, Sexual Transmitted

Diseases (STD) in women

GUM Genito-Urinary Males + Prostatitis, epididymo-orchitis, STD in

men

No

defined

site (NDS)

BAC Bacteraemia with no clear anatomic site and no shock

SEPSIS Sepsis, sepsis syndrome or septic shock with no clear anatomic

site

Malaria

HIV Human immunodeficiency virus

PUO Pyrexia of Unknown Origin - Fever syndrome with no identified

source or site of infection

PUO-HO Fever syndrome in the non-neutropaenic Haematology–

Oncolgy patient with no

identified source of pathogen

FN Fever in the Neutropenic patient

LYMPH Infection of the lymphatics as the primary source of infection

e.g. suppurative lymphadenitis

conditions

antibiotics

not required

Antibiotic prescribed with documentation for which there is no

above diagnosis group

MP-GEN Drug is used as Medical Prophylaxis in general, without

targeting a specific site, e.g.

antifungal prophylaxis during immunosuppression

PROK Antimicrobial (e.g. erythromycin) prescribed for Prokinetic use

Neonatal MP-MAT Drug is used as Medical Prophylaxis for MATERNAL risk factors

e.g. maternal prolonged rupture of membranes

NEO-MP Drug is used as Medical Prophylaxis for NEONATE risk factors

e.g. VLBW (Very Low

Birth Weight) and IUGR (Intrauterine Growth Restriction)

272

Explanatory Notes III – Patient Form

Adapted from the Global Point Prevalence Survey (2017 GLOBAL-PPS): http://www.global-pps.com/documents

1) Combinations of an antibiotic and an enzyme inhibitor

a) Ampicillin and enzyme inhibitor: report only ampicillin dose

(J01CR01) b) Amoxicillin and enzyme inhibitor: report only amoxicillin dose

(J01CR02) c) Ticarcillin and enzyme inhibitor: report only ticarcillin dose

(J01CR03) d) Piperacillin and enzyme inhibitor: report only piperacillin dose

(J01CR05) e) Imipenem and enzyme inhibitor: report only imipenem dose

(J01DH51) f) Panipenem and betamipron: report only panipenem (J01DH55)

Example:

• Augmentin® 1.2g IV: 1g (amoxicillin) + 200mg (clavulanic acid), report only 1 g

• Piperacillin® 4.5g IV: 4g (piperacillin) + 500mg (tazobactam), report only 4 g

2) Other combinations of multiple antimicrobial substances:

a) J01EE01 Sulfamethoxazole and Trimethoprim: report the total

amount of sulfamethoxazole and trimethoprim Example:

• Co-trimoxazole 960mg: (sulfamethoxazole. 800mg + trimethoprim 160mg), report 960mg

Further information on agents included for the Global-PPS is available in the antimicrobial list. Only antimicrobial substance name need to be written down, NOT the ATC codes (excel file - available at website under documents: Global-PPS_antimicrobial_list.xlsx) http://www.global-pps.com/

273

Explanatory Notes IV – Patient Form Adapted from the Global Point Prevalence Survey (2017 GLOBAL-PPS): http://www.global-

pps.com/documents

(Type of Indications: Select one for each antibiotic or other antimicrobial)

CAI: Community

acquired infection

CAI Symptoms start <48 hours from admission

to hospital (or present on admission).

HAI: Healthcare

associated infection

- Symptoms start ≥ 48 hours after admission

to hospital.

HAI1 Post-operative surgical site infection (within

30 days of surgery OR; 1 year after implant

surgery)

HAI2 Intervention related infections including CR-

BSI, VAP and C- UTI

HAI3 C. difficile associated diarrhoea (CDAD) (>48

h post- admission or <30 days after

discharge from previous admission

episode).

HAI4 Other hospital acquired infection (includes

HAP, etc)

HAI5 Infection present on admission from

another hospital (patient with infection

from another hospital)

HAI6 Infection present on admission from long-

term care facility (LTCF) or Nursing Home*.

SP: Surgical

prophylaxis

SP1 Single dose

SP2 One day

SP3 >1 day

MP: Medical

prophylaxis

MP For example, long term use to prevent UTI's

or use of antifungals in patients undergoing

274

chemotherapy or penicillin in asplenic

patients, etc.

OTH: Other OTH For example, erythromycin as a motility

agent (motilin agonist).

UNK: Unknown UNK Completely unknown indication

CR-BSI = Catheter Related-Blood Stream Infection; C-UTI = Catheter related-Urinary

Tract Infection; HAP = Hospital Acquired Pneumonia; VAP = Ventilator Associated

Pneumonia.

* Long-term care facilities represent a heterogeneous group of healthcare facilities,

with care ranging from social to medical care. These are places of collective living where

care and accommodation is provided as a package by a public-agency, non-profit or

private company (e.g. nursing homes, residential homes).

Supplement 4

Antibiotic consumption survey standard operating procedures and data

collection tools The Antimicrobial Consumption Survey – Study Details

Antimicrobial Consumption Survey

We will conduct a survey of antibiotic prescribing in relation to clinical diagnosis

in all hospitalized adults, children and neonates across 14 Kenyan hospitals.

The survey will collate data on drug, dose and indicators of antibiotic (and other

antimicrobial) prescribing across inpatients in the participant facilities, with the

aim of identifying targets to improving quality of prescribing.

Relevant data is generally held in clinical patient notes and treatment sheets;

the survey team will therefore work with medical record administrators and

clinical staff at each ward to collate relevant information.

The survey will involve reviewing all prescriptions of antibiotics and other

antimicrobials among hospitalized individuals, by inspecting treatment sheets,

and establishing the diagnosis and relevant investigations that would support

the prescription, by inspecting all other relevant clinical patient notes.

275

Participant hospitals are committed to facilitate the task of the survey team by

providing:

Access to - and support from - relevant ward staff,

Access to relevant patient notes,

Access to a room with tables, chairs and electricity.

The ACS will follow standard operating procedures and data capture forms

currently in use as part of the 'Global Point Prevalence Survey on Antimicrobial

Consumption and Resistance (Global PPS)' and will consequently collate data on

additional antimicrobials relevant to Global-PPS activities (i.e. antifungals,

antivirals, antimalarials). Data capture forms and details of how the survey will

be conducted are given in the following sections. The Global Point Prevalence

Survey (PPS) – Year 2017 protocol can be found at http://www.global-

pps.com/documents/

1. Outline Protocol

Key aspects of the protocol are described below:

1) Preparation of a hospital ward list;

2) What guidelines are being used to guide prescription of antibiotics and

other antimicrobials based on diagnosis or clinical condition. This

information may be available at the facility-level or at the ward-level.

3) Inclusion of all eligible hospital wards in the survey (except for the case

of Kenyatta National Hospital, where only neonatal wards will be

included in the hospital);

4) Survey of each ward only once on a single day. Different wards may be

surveyed on different dates;

5) Inclusion criteria:

a. All inpatients admitted on the ward before 8 o'clock in the

morning on the day of the survey, including all neonate healthy

infants in the case of maternity wards (denominator data);

b. All inpatients "on antibiotics or other eligible antimicrobials",

and who were admitted before 8 o'clock in the morning on the

day of the survey (numerator data). Patients "on antibiotics"

are those undergoing antibiotic treatment even if the drug is

not administered the day of the survey. The definition also

includes patients who received one or more doses of

276

antibiotics or other eligible anti-infective agents intended as

surgical prophylaxis in the 24 h prior to 8:00 am on the day of

the survey.

6) Exclusion criteria:

a. All day admissions and outpatients (denominator and

numerator). This includes all healthy mothers and babies in a

maternity ward, who will be discharged within 24 hours of

admission, following successful delivery. These patients should

be regarded as day patients even if present in the ward at the

time of the survey;

b. Patients discharged before or admitted after 8 o'clock in the

morning on the day of the survey (denominator and

numerator);

c. Inpatients on antibiotics or other eligible antimicrobials, where

treatment was no longer active/ongoing at 8 o'clock in the

morning on the day of the survey (numerator).

d. Amenity, Psychiatry, ENT, Renal, ICU, NICU (including in

Kenyatta hospital), Neurosurgery and Eye wards are excluded

from the survey, as these units are not present in all hospitals.

7) Denominator data:

a. Total number of admitted eligible patients (i.e. admitted before

8 am) in the ward surveyed;

b. Total number of available beds attributed to inpatients at 8 am

of the ward surveyed (i.e. number of total inpatient beds

[occupied plus empty] at the time of the survey).

8) Antibiotics and other eligible antimicrobial agents: Classification of

antibiotics and other antimicrobials will agree with the ATC/DDD Index

2017 of the WHO Collaborating Centre for Drug Statistics Methodology;

a. Inclusion criteria - Anti-infective agents within the following

ATC codes:

i. J01: Antibiotics for systemic use

ii. A07AA: Antibiotics used as intestinal anti-infectives

iii. P01AB: Antiprotozoals used as antibacterial agents,

nitroimidazole derivatives

iv. J04A: Antibiotics and other drugs used for treatment of

tuberculosis

277

v. J05AH: Antivirals used for influenza - Neuraminidase

inhibitors

vi. J02 and D01BA: Antimycotics and antifungals for

systemic use

vii. P01B: Antimalarials

b. Exclusion criteria: Antibiotics and other antimicrobials for

topical use.

9) Completion of data collection forms:

a. Ward form (denominator)

b. Patient form (numerator)

10) To complete 'Patient forms', the investigators will review all [medical,

nursing and drug prescription chart] patient records. If the information

available is not sufficient surveyor/s may request additional

information from nurses, pharmacists or doctors caring for the patient.

Searching for information from other sources such a laboratory

computer systems, phoning laboratories etc., is not required. At no

point shall there be any discussion about the appropriateness (or lack

thereof) of the prescribed medication. The ward staff MUST NOT feel

evaluated at the individual level.

11) Confidentiality:

a. Data stored for analysis WILL NOT bear the names of patients

or staff and neither patients, staff nor hospitals will be named

in any published reports. Inpatient numbers (i.e. Patient IP

number) will be removed from the database prior to analysis

and following data cleaning. It is not permissible to discuss the

information collected with anyone outside the research team.

12) Data entry conventions:

a. For numerical fields, enter -1 if the information is missing

(applicable but not available).

b. For date fields, enter 01/01/1913 if the information is missing

(applicable but not available)

c. For date fields, enter 01/01/1914 if not applicable

d. For drop down lists with the option 'empty', choosing 'empty'

means that the information requested is not available.

13) Data Entry and Synchronization

a. After data entry into REDCap, the data clerks will be expected

to synchronize the data daily. This is to ensure the data are

278

stored at the KEMRI/Wellcome Trust servers. You will be

provided with internet modems to allow synchronization of the

data over the internet.

279

Supplement 5: Table of available guidelines by disease Condition Adult/

Paediatric

Treatment Options Source Reference

1 Pneumonia Adult fluoroquinolone (moxifloxacin,

gemifloxacin, or levofloxacin

[750 mg])

Infectious Diseases

Society of America;

ATS: American

Thoracic Society

2019

Metlay JP, Waterer GW, Long AC, Anzueto A, Brozek J, Crothers K,

Cooley LA, Dean NC, Fine MJ, Flanders SA, Griffin MR. Diagnosis and

Treatment of Adults with Community-acquired Pneumonia. An Official

Clinical Practice Guideline of the American Thoracic Society and

Infectious Diseases Society of America. American Journal of

Respiratory and Critical Care Medicine. 2019 Oct 1;200(7):e45-67.

An antipneumococcal beta-

lactam (preferred agents:

cefotaxime, ceftriaxone, or

ampicillin-sulbactam; or

ertapenem for selected

patients) PLUS a macrolide

(azithromycin, clarithromycin, or

erythromycin)

Treat with benzyl penicillin 2

mega units IM IV 6 hourly +

gentamicin 240 mg IM IV once a

day 5 days OR IV ceftriaxone 2g

every 24 hours OR erythromycin

500mg 6 hourly for 5 days

Clinical

Management and

Referral Guidelines

– Volume III:

Clinical Guidelines

for Management

and Referral of

Common

Conditions at

Levels 4–6:

Hospitals.

Ministry of Medical Services and Ministry of Public Health and

Sanitation. Clinical Management and Referral Guidelines – Volume III:

Clinical Guidelines for Management and Referral of Common

Conditions at Levels 4–6: Hospitals. Nairobi: Ministry of Medical

Services and Ministry of Public Health and Sanitation; 2009.

Amoxicillin 500 mg orally three

times daily

2009 guideline

recommendations

Lim WS, Baudouin SV, George RC, et al. BTS guidelines for the

management of community-acquired pneumonia in adults: update

280

Amoxicillin 500 mg to 1 gram

orally three times

daily plus clarithromycin 500 mg

orally twice daily

of the British

Thoracic Society

2009. Thorax 2009; 64 Suppl 3:iii1. Copyright © 2009 BMJ Publishing

Group Ltd.

Doxycycline 200 mg loading dose

then 100 mg

orally or levofloxacin 500 mg

orally once daily or moxifloxacin

400 mg orally once daily

amoxicillin 500 mg IV three

times daily or benzylpenicillin

(penicillin G) 1.2 grams IV four

times daily plus clarithromycin

500 mg IV twice daily

Adult Amoxicillin/Clavulanic Acid/

cefuroxime or Ceftriaxone and

Macrolide

Kenyatta National

Hospital Kenya

The KNH Guide to Empiric Antimicrobial Therapy 2018

Children benzylpenicillin + gentamicin Basic Paediatric

Protocols for ages

up to 5 years.

2016

Ministry of Health. Basic Paediatric Protocols for ages up to 5 years.

2016 [Available from: http://idoc-

africa.org/images/documents/2016/Basic_Paediatric_Protocol_2016/

MAY%2023rd%20BPP%202016%20SA.pdf.

2 Skin and Soft

Tissue

Infections

Adult/

Children

Cloxacillin Clinical

Management and

Referral Guidelines

– Volume III:

Clinical Guidelines

for Management

and Referral of

Ministry of Medical Services and Ministry of Public Health and

Sanitation. Clinical Management and Referral Guidelines – Volume III:

Clinical Guidelines for Management and Referral of Common

Conditions at Levels 4–6: Hospitals. Nairobi: Ministry of Medical

Services and Ministry of Public Health and Sanitation; 2009.

281

Common

Conditions at

Levels 4–6:

Hospitals.

Adult Amoxicillin/Clavulanic Acid or

Clindamycin or Doxycycline

Kenyatta National

Hospital Kenya

The KNH Guide to Empiric Antimicrobial Therapy 2018

Adult

(extended

hospitalizat

ion with

invasive

procedure

Piperacillin/Tazobactam+

Amikacin or Cefipime +Amikacin

Adult/Child

ren

Vancomycin +piperacillin Infectious Diseases

Society of America

Dennis L. Stevens, Alan L. Bisno, Henry F. Chambers, E. Patchen

Dellinger, Ellie J. C. Goldstein, Sherwood L. Gorbach, Jan V.

Hirschmann, Sheldon L. Kaplan, Jose G. Montoya, James C. Wade,

Practice Guidelines for the Diagnosis and Management of Skin and

Soft Tissue Infections: 2014 Update by the Infectious Diseases Society

of America, Clinical Infectious Diseases, Volume 59, Issue 2, 15 July

2014, Pages e10–e52, https://doi.org/10.1093/cid/ciu296

Penicillin, Ceftriaxone,

Clindamycin, Cephazolin

Ampicillin - Sulbactam

Ceftriaxone+Metronidazole

Ciprofloxacin+Metronidazole

Levofloxacin+Metronidazole

282

National Institute

for Health and

Care Excellence

(NICE)

Guideline on cellulitis and erysipelas – Antimicrobial

prescribing (2019)--

https://www.nice.org.uk/guidance/ng141/resources/visual-summary-

pdf-6908401837

3 Urinary Tract

Infections

Adult-

Lower UTI

Cotrimoxazole Clinical

Management and

Referral Guidelines

– Volume III:

Clinical Guidelines

for Management

and Referral of

Common

Conditions at

Levels 4–6

Adult -

Upper UTI

Gentamicin, Ciprofloxacin Ministry of Medical Services and Ministry of Public Health and

Sanitation. Clinical Management and Referral Guidelines – Volume III:

Clinical Guidelines for Management and Referral of Common

Conditions at Levels 4–6: Hospitals. Nairobi: Ministry of Medical

Services and Ministry of Public Health and Sanitation; 2009.

Adult(

Category 1)

Nitrofurantoin or Cefuroxime or

Ciprofloxacin

Kenyatta National

Hospital Kenya

The KNH Guide to Empiric Antimicrobial Therapy 2018

Adult

(Category

2)

Nitrofurantoin or Etrapenem or

Piperacillin/Tazobactam

Adult

(Category

3)

Meropenem or Imipenem+

Amikacin or

Piperacillin/Tazobactam+Amikaci

n

Adult Ceftriaxone Infectious Diseases

Society of

America/NICE

Thomas M. Hooton, Suzanne F. Bradley, Diana D. Cardenas, Richard

Colgan, Suzanne E. Geerlings, James C. Rice, Sanjay Saint, Anthony J.

Schaeffer, Paul A. Tambayh, Peter Tenke, Lindsay E. Nicolle, Diagnosis,

Prevention, and Treatment of Catheter-Associated Urinary Tract Ciprofloxacin

Levofloxacin

283

Trimethoprim-Sulfamethoxazole Infection in Adults: 2009 International Clinical Practice Guidelines from

the Infectious Diseases Society of America, Clinical Infectious Diseases,

Volume 50, Issue 5, 1 March 2010, Pages 625–663,

https://doi.org/10.1086/650482

Amoxicillin Clavulanate

Children Co-amoxiclav Cefuroxime

Ceftriaxone Gentamicin

Amikacin

National Institute

for Health and

Care Excellence

(NICE)

Pyelonephritis (acute): antimicrobial prescribing

https://www.nice.org.uk/guidance/ng111/resources/visual-summary-

pdf-6544161037 UTI (catheter): antimicrobial prescribing

Adult Ceftriaxone, Ciprofloxacin,

Levofloxacin,

National Institute

for Health and

Care Excellence

(NICE)

UTI (catheter): antimicrobial prescribing

https://www.nice.org.uk/guidance/ng113/resources/visual-summary-

pdf-6599495053

Lower UTI Nitrofurantoin National Institute

for Health and

Care Excellence

(NICE)

UTI (lower): antimicrobial prescribing

https://www.nice.org.uk/guidance/ng109/resources/visual-summary-

pdf-6544021069

4 Intrabdominal

Infections

Adults

(Category

1)

Ceftriaxone+Metronidazole or

Ciprofloxacin+Metronidazole or

Tigecycline+ Metronidazole

Kenyatta National

Hospital Kenya

The KNH Guide to Empiric Antimicrobial Therapy 2018

Adults(Cate

gory 2)

Etrpenem or

Piperacillin/Tazobactam+

Aminoglycoside or

Ceftazidime+Aminoglycoside+

Metronidazole or Tigecycline+

Metronidazole

Adults(Cate

gory 2)

Imipenem+/-Aminoglycoside or

Meropenem +/- aminoglycoside

284

or cefipime+

aminoglycoside+metronidazole

Children Ceftriaxone, cefotaxime,

cefepime, or ceftazidime, each in

combination with

metronidazole; gentamicin or

tobramycin, each in combination

with metronidazole or

clindamycin, and with or without

ampicillin

Infectious Diseases

Society of America

Joseph S. Solomkin, John E. Mazuski, John S. Bradley, Keith A Rodvold,

Ellie J.C. Goldstein, Ellen J. Baron, Patrick J. O'Neill, Anthony W. Chow,

E. Patchen Dellinger, Soumitra R. Eachempati, Sherwood Gorbach,

Mary Hilfiker, Addison K. May, Avery B. Nathens, Robert G. Sawyer,

John G. Bartlett, Diagnosis and Management of Complicated Intra-

abdominal Infection in Adults and Children: Guidelines by the Surgical

Infection Society and the Infectious Diseases Society of America,

Clinical Infectious Diseases, Volume 50, Issue 2, 15 January 2010,

Pages 133–164, https://doi.org/10.1086/649554 Adults Cefazolin, cefuroxime,

ceftriaxone, cefotaxime,

ciprofloxacin, or levofloxacin,

each in combination with

metronidazole

5 CNS

Infections

Adults Benzylpenicillin 4 mega units IV 6

hourly for 14–

21 days OR chloramphenicol 1g

IV 6 hourly for 14 days, OR

ceftriaxone 24g/

day IV 12 hourly for 14–21 days,

Vancomycin 2g/day IV 8–12

hourly OR

meropenem 2g/day IV 8 hourly

Clinical

Management and

Referral Guidelines

– Volume III:

Clinical Guidelines

for Management

and Referral of

Common

Conditions at

Levels 4–6:

Ministry of Medical Services and Ministry of Public Health and

Sanitation. Clinical Management and Referral Guidelines – Volume III:

Clinical Guidelines for Management and Referral of Common

Conditions at Levels 4–6: Hospitals. Nairobi: Ministry of Medical

Services and Ministry of Public Health and Sanitation,; 2009.

285

Children Ceftriaxone Basic Paediatric

Protocols for ages

up to 5 years.

2016

Ministry of Health. Basic Paediatric Protocols for ages up to 5 years.

2016 [Available from: http://idoc-

africa.org/images/documents/2016/Basic_Paediatric_Protocol_2016/

MAY%2023rd%20BPP%202016%20SA.pdf.

Adults Vancomycin plus ampicillin plus

a third-generation cephalosporin

Infectious Disease

Society of America

Liu C, Bayer A, Cosgrove SE, et al. Clinical Practice Guidelines by the

Infectious Diseases Society of America for the Treatment of

Methicillin-Resistant Staphylococcus Aureus Infections in Adults and

Children: Executive Summary. Clin Infect Dis 2011; 52:285. Adults >50 Vancomycin plus ampicillin plus

a third-generation cephalosporin

Adults Post

Trauma

Vancomycin plus cefepime; OR

vancomycin plus ceftazidime; OR

vancomycin plus meropenem

6 Sepsis/Septic

Shock

Adults vancomycin with one of the

following:

3rd generation (e.g., ceftriaxone

or cefotaxime) or 4th generation

cephalosporin (cefepime), or

beta-lactam/beta-lactamase

inhibitor (eg, piperacillin-

tazobactam, ticarcillin-

clavulanate), or

carbapenem (e.g., imipenem or

meropenem)

Surviving Sepsis

Campaign

Rhodes A, Evans LE, Alhazzani W, Levy MM, Antonelli M, Ferrer R, et

al. Surviving Sepsis Campaign: International Guidelines for

Management of Sepsis and Septic Shock: 2016. Intensive Care

Medicine. 2017;43(3):304-77.

286

Children Vancomycin (15 mg/kg,

maximum 1 to 2 g, for the initial

dose) PLUS cefotaxime (100

mg/kg, maximum 2 g, for the

initial dose) OR ceftriaxone (75

mg/kg, maximum 2 g, for the

initial dose)

•Consider adding an

aminoglycoside (e.g.,

gentamicin) for possible GU

source and/or piperacillin with

tazobactam, clindamycin or

metronidazole for possible GI

source

Surviving Sepsis

Campaign

Rhodes A, Evans LE, Alhazzani W, et al. Surviving Sepsis Campaign:

International Guidelines for Management of Sepsis and Septic Shock:

2016. Intensive Care Med 2017; 43:304.

7 Bone and

Joint Infection

Adults/Chil

dren

Clindamycin( Adults),Cloxacillin +

chloramphenicol

Cloxacillin + gentamicin or

Amoxicillin + gentamicin for

septic arthritis

Clinical

Management and

Referral Guidelines

– Volume III:

Clinical Guidelines

for Management

and Referral of

Common

Conditions at

Levels 4–6:

Ministry of Medical Services and Ministry of Public Health and

Sanitation. Clinical Management and Referral Guidelines – Volume III:

Clinical Guidelines for Management and Referral of Common

Conditions at Levels 4–6: Hospitals. Nairobi: Ministry of Medical

Services and Ministry of Public Health and Sanitation; 2009.

287

Adults Vancomycin PLUS a third- or

fourth-generation cephalosporin

(such as ceftriaxone,

ceftazidime, or cefepime)

Infectious Disease

Society of America

Mader JT, Cantrell JS, Calhoun J. Oral ciprofloxacin compared with

standard parenteral antibiotic therapy for chronic osteomyelitis in

adults. J Bone Joint Surg Am 1990; 72:104.Davey PG, Rowley DR,

Phillips GA. Teicoplanin--home therapy for prosthetic joint infections.

Eur J Surg Suppl 1992; :23.Lazzarini L, Lipsky BA, Mader JT. Antibiotic

treatment of osteomyelitis: what have we learned from 30 years of

clinical trials? Int J Infect Dis 2005; 9:127.Black J, Hunt TL, Godley PJ,

Matthew E. Oral antimicrobial therapy for adults with osteomyelitis or

septic arthritis. J Infect Dis 1987; 155:968.Gentry LO, Rodriguez-Gomez

G. Ofloxacin versus parenteral therapy for chronic osteomyelitis.

Antimicrob Agents Chemother 1991; 35:538.Lew DP, Waldvogel FA.

Quinolones and osteomyelitis: state-of-the-art. Drugs 1995; 49 Suppl

2:100.

Children Vancomycin PLUS Nafcillin or

Ofloxacin PLUS a third- or

fourth-generation cephalosporin

(such as ceftriaxone,

ceftazidime, or cefepime) and

Clindamycin

8 Post

Caesarean

Section

Prophylaxis

Adults First Generation Cephalosporin

or Add metronidazole,

azithromycin, or doxycycline to a

first- or second-generation

cephalosporin

Infectious Disease

Society of America

Dale W. Bratzler, E. Patchen Dellinger, Keith M. Olsen, Trish M. Perl,

Paul G. Auwaerter, Maureen K. Bolon, Douglas N. Fish, Lena M.

Napolitano, Robert G. Sawyer, Douglas Slain, James P. Steinberg,

Robert A. Weinstein, Clinical practice guidelines for antimicrobial

prophylaxis in surgery, American Journal of Health-System Pharmacy,

Volume 70, Issue 3, 1 February 2013, Pages 195–283,

https://doi.org/10.2146/ajhp120568

9 Prophylaxis

for bone and

joint surgery

Adults Cephazolin, Clindamycin,

vancomycin

Infectious Disease

Society of America

Dale W. Bratzler, E. Patchen Dellinger, Keith M. Olsen, Trish M. Perl,

Paul G. Auwaerter, Maureen K. Bolon, Douglas N. Fish, Lena M.

Napolitano, Robert G. Sawyer, Douglas Slain, James P. Steinberg,

Robert A. Weinstein, Clinical practice guidelines for antimicrobial

prophylaxis in surgery, American Journal of Health-System Pharmacy,

Volume 70, Issue 3, 1 February 2013, Pages 195–283,

https://doi.org/10.2146/ajhp120568

288

10 Prophylaxis

for GIT

surgery

Adults Cefazolin, cefoxitin, cefotetan,

ceftriaxone,ampicillin–

sulbactam, Clindamycin or

vancomycin +aminoglycoside or

aztreonam orfluoroquinolone or

Metronidazole + aminoglycoside

or fluoroquinolone

Infectious Disease

Society of America

Dale W. Bratzler, E. Patchen Dellinger, Keith M. Olsen, Trish M. Perl,

Paul G. Auwaerter, Maureen K. Bolon, Douglas N. Fish, Lena M.

Napolitano, Robert G. Sawyer, Douglas Slain, James P. Steinberg,

Robert A. Weinstein, Clinical practice guidelines for antimicrobial

prophylaxis in surgery, American Journal of Health-System Pharmacy,

Volume 70, Issue 3, 1 February 2013, Pages 195–283,

https://doi.org/10.2146/ajhp120568

11 Prophylaxis

for Urological

Surgery

Adults Cefazolin ± aminoglycoside,

cefazolin ± aztreonam,

ampicillin–sulbactam,

Fluoroquinolone, trimethoprim–

sulfamethoxazole, Cefazolin +

metronidazole, cefoxitin,

aminoglycoside +

metronidazole or clindamycin

Infectious Disease

Society of America

Dale W. Bratzler, E. Patchen Dellinger, Keith M. Olsen, Trish M. Perl,

Paul G. Auwaerter, Maureen K. Bolon, Douglas N. Fish, Lena M.

Napolitano, Robert G. Sawyer, Douglas Slain, James P. Steinberg,

Robert A. Weinstein, Clinical practice guidelines for antimicrobial

prophylaxis in surgery, American Journal of Health-System Pharmacy,

Volume 70, Issue 3, 1 February 2013, Pages 195–283,

https://doi.org/10.2146/ajhp120568

12 Neonatal

Infections

Neonates Penicillin+ Gentamicin,

Ceftazidime, Metronidazole,

Ceftriaxone, Cefotaxime

Basic Paediatric

Protocols for ages

up to 5 years.

2016

Ministry of Health. Basic Paediatric Protocols for ages up to 5 years.

2016 [Available from: http://idoc-

africa.org/images/documents/2016/Basic_Paediatric_Protocol_2016/

MAY%2023rd%20BPP%202016%20SA.pdf.

13 Genitourinary

system

Ministry of Health

Kenya

Ministry of Health. Kenya National Guidelines for Prevention,

Management and Control of Sexually Transmitted Infections. Nairobi:

National AIDS and STI Control Programme; 2018.

Male STD IM ceftriaxone/Gentamicin and

Azithromycin

Cervicitis IM ceftriaxone/Gentamicin and

Azithromycin

Pelvic

Inflammato

ry Disease

Cefixime/Ceftriaxone/Gentamici

n and Doxycline and

Metronidazole

289

Supplement 6: Heat map showing the physical availability of guidelines by disease condition

across the departments

290

291

Chapter 7

Using treatment guidelines to improve

antibiotic use: Insights from an antibiotic

point prevalence survey in Kenyan public

hospitals

Michuki Maina, Jacob McKnight, Olga Tosas- Auguet,

Constance Schultsz and Mike English

In Press ( BMJ Global Health )

292

Improving antibiotic use in resource-limited regions; the

role of treatment guidelines

Key Words: Antimicrobial Resistance, Antibiotics, Treatment guidelines

Summary Box

• Clinical practice guidelines have the potential to improve quality of care

through improving decision making and antibiotic prescription. These

guidelines are particularly important in areas with limited laboratory

and specialist capacity.

• For some of the common conditions managed in the hospitals,

guidelines are not available or are outdated.

• To reduce irrational antibiotic use and contain the threat of

antimicrobial resistance, the process of guideline development should

prioritise the most common diseases.

• The process of developing context-appropriate clinical guidelines

requires input from all relevant stakeholders with leadership from the

Ministry of Health. This process needs to have a clear plan for

dissemination, training and future updates.

293

Background

Antimicrobial resistance (AMR) is a significant public health threat that is

expected to worsen as more drug-resistant organisms emerge [1]. This situation

is further exacerbated by the low rate of discovery of new antimicrobial agents

that could act against drug-resistant microorganisms. AMR could retard

economic growth in low-income countries and delay attainment of the

sustainable development goals [2].

There are multiple drivers of AMR, but one of the key drivers has been the

irrational use of antimicrobial agents [3]. In hospitals, lack of timely and

accurate diagnostic tests, including microbiology for bacterial speciation and

drug susceptibility testing, leads to unnecessary antimicrobial use, fueling

resistance and health care costs [4]. While countries must work to improve

diagnostic capabilities and increase laboratory capacity to enhance diagnostic

accuracy, it is also important to complement this new capacity with locally

relevant guidelines. Providing context-specific, applicable, and regularly

updated treatment guidelines to frontline doctors is an effective means to

improve antibiotic usage and clinical care [5].

Clinical practice guidelines (CPG) provide a standardised and systematic

approach to responding to disease, including the treatment. The guidelines are,

however, more effective in the context of a functioning health system with

adequate clinicians, drugs, diagnostics and a supportive environment to the

clinicians and patients[6]. This multifaceted approach of improving

antimicrobial usage is illustrated in Figure 1.

294

Figure 1 Improving antimicrobial usage requires clinicians, laboratory and radiological support in

an environment that provides treatment guidelines, training, feedback and patient

communication.

Here, we present data from an antibiotic point prevalence survey that

illustrates the continued importance of developing guidelines for improving

diagnosis and treatment. We propose that guidelines should be developed to

target common diseases in limited-resource settings as a priority and lastly, we

illustrate how guideline development and dissemination at scale can be

achieved using the example of Kenya’s Basic Paediatric Protocols (BPP).

295

Using guidelines to improve diagnosis and antibiotic use

A point prevalence survey was conducted across 14 public county hospitals

(formerly district hospitals) in Kenya. These hospitals, with varying bed

capacities, are located in urban and rural areas of Kenya with high and low

malaria endemicity. They provide multi-speciality inpatient care, which

includes; maternal, neonatal, adult and paediatric medical and surgical units.

Data from this survey conducted among hospitalised patients revealed that

large proportions of patients received antibiotic treatment for conditions that

did not warrant antibiotics [7]. However, this inappropriate use of antibiotics

was much less prevalent in the paediatric medical (14% [36/261) and neonatal

units (5% [11/224]) where treatment guidelines were physically available than

on adult medical units where treatment guidelines were absent, and 33%

[140/421] of the patients were inappropriately treated.

In addition to lack of guidelines, the literature on irrational antibiotic use

suggests that the level of training of the prescribers, the fear of the clinicians

missing an existing infection, fear of lawsuits, fear of being reprimanded by

more senior colleagues, and pressures from patients, nurses and other ward

staff can especially encourage overtreatment [8] [9]. Availing standard

guidelines can help address many of these factors and give confidence to the

clinicians on what comprises best practice while also improving the accuracy of

diagnoses and treatment [10].

The need for Guidelines

While it may be ideal to have guidelines for all diseases, immediate efforts

should be directed to the most common ailments. On the neonatal and

paediatric units we examined, the clinical practice guidelines available in the

296

wards [11] provided treatment advice that spanned 94% (211/224) and 55%

(143/261) of the admissions, respectively. There were guidelines for common

conditions in adults or on surgical units.

The main adult conditions for which antibiotics were prescribed among the

adult medical and surgical populations included; pneumonia, obstetric and

gynaecological infections, HIV associated infections, central nervous system

(CNS) infections, skin and soft tissue infections (SSTI), and antibiotics for

surgical prophylaxis. In the paediatric medical unit, the common conditions

included; pneumonia, CNS infections, gastrointestinal infections and sepsis [7].

Here we use the example of pneumonia and CNS infections in adults and SSTI

in surgical units to illustrate the need for guidelines.

In the adult medical wards, pneumonia and CNS infections accounted for 22%

[94/421] and 20% [50/421] of admissions, respectively. Based on predefined

criteria, 26% and 28% of these patients admitted with pneumonia and CNS

infections, respectively, received inappropriate treatment. In the surgical units,

SSTI which lack current local guidelines were a common cause of hospitalisation

in adults (25% [135/543]) and children (60% [32/53]). Of these patients with

SSTI, the documented antibiotic treatment was inappropriate in 69% and 43%

of the adults and children, respectively. The choice of antibiotics used to treat

these skin infections varied widely across the hospitals surveyed. Additionally,

there was a significantly higher use of nitroimidazole derivatives compared to

the preferred beta-lactam antibiotics [12].

Availing approved guidelines for these conditions could ensure their treatment

is standardised across hospitals. To reduce AMR, these guidelines should be in

line with the recommendations by the WHO essential medicines list that

encourages the use of the access group drugs (generally having a narrow

297

spectrum of activity) as first and second-line therapy under the Access, Watch

and Reserve (AWaRe) categorisation [13]. Treatment guidelines for adults in

medical and surgical units which cover these, and other common conditions

were developed in Kenya. However, they have not been updated for 12 years;

they were not disseminated in easy to use formats or at scale and were not

found in any of the hospitals visited [14]. Therefore, it is essential to update or

replace these guidelines in a format that can be widely and rapidly

disseminated. It would be advisable to draw on international guidelines,

including those from WHO [15-17] and include context-specific modifications

[18]. This process should be driven from the “bottom-up” by the professional

medical and surgical associations with strategic direction offered by the

Ministry of Health as the Ministry is mandated to generate health policies.

The potential for national guidelines

Unlike in high-income settings where guidelines may be hospital or region-

specific, in many low- and middle-income countries, guidelines tend to be for

national use, including the private sector. Key to the development process,

therefore, are the Ministries of Health who provide direction on what

guidelines need to be developed based on the local disease burden. The

Ministry also plays a role in ensuring adequate representation by all relevant

stakeholders in the development process that should be transparent and not

undermined by conflicts of interest or undue influence from specific individuals

or groups (including funders) [19]. Guideline development should specifically

extend to include plans for training and dissemination activities targeting all

relevant clinicians in pre-service training or practice in the public, not-for-profit

and private sectors. This includes availing the guidelines in electronic versions

for higher utilisation [20].

298

Other stakeholders, especially organisations such as WHO can play a vital role

by providing access to high-quality evidence syntheses and offering training and

technical support to the process of guideline development [21].

The Kenyan Basic Paediatric Protocols as an example of a national

guideline development process

Developing guidelines is a multidisciplinary effort that requires input from

expert clinicians, representatives from professional bodies and end-users,

economists and methodologists. These teams need to be responding to

relevant clinical needs that require the development of guidelines [21].

In Kenya, this approach was adopted to develop the Basic Paediatric Protocols

(BPP) first published in 2006. The development of these guidelines was initiated

and overseen by the Ministry of Health with input from the professional

paediatric association, universities and clinicians [22]. As the process evolved,

topics were identified in response to clinician queries for conditions that lacked

clear guidelines, and systematic reviews were conducted to generate context-

appropriate evidence. This evidence was presented to multiple stakeholders,

and guidelines were then developed through consensus. The BPP underwent

three updates in 2010, 2013 and 2016 to include new and emerging evidence

[11, 23].

Guideline availability, however, does not necessarily translate into their use. It

is, therefore, essential to include a plan for training and implementation as part

of the guideline development process [23]. The development of the BPP was

accompanied by the roll-out of the BPP-based Emergency Triage Assessment

and Treatment Plus Admission (ETAT+) training to clinicians in public and

private hospitals and medical students. Many thousands of health workers and

students have been trained since the guidelines were first rolled out, and

299

numerous low-cost, short guidelines booklets have been distributed [23]. Since

their introduction, better case definition and management of pneumonia and

diarrhoeal diseases, improved clinical documentation and a decline in the use

of inappropriate cough medications have been reported amongst other

benefits [10, 24, 25].

Conclusion

Availing up to date treatment guidelines to clinicians provides an opportunity

to reduce inappropriate antibiotic use in hospitals. There are apparent gaps in

guideline development and availability, especially for common adult medical

conditions and across all ages in surgical care. Guideline development can be

used to build consensus across a broad spectrum of the clinical community on

contextually appropriate treatments. Their development must be accompanied

by clear and adequate dissemination strategies to ensure all the clinicians

making decisions daily, understand the rationale for the recommended strategy

and have access to guidance at the point of care.

Authors’ contributions

The roles of the contributors were as follows: M.M, J.M, O.T, C.S and M.E

conceived the study. M.M conducted surveys and data collection. O.T and ME

assisted M.M in analysis and interpretation of the data. M.M, J.M, O.T, C.S and

M.E drafted and critically revised the manuscript for intellectual content. All

authors read and approved the final manuscript.

300

Acknowledgements

The authors would like to thank all the hospitals that participated in the initial

surveys and follow up interviews. This work is published with the permission of

the director of KEMRI.

Funding

This work was supported by funds from the economic and social research

council ESRCS # ES/P004938/1, and a Senior Research Fellowship awarded to

M.E by The Wellcome Trust (#207522). M.M is supported by a grant from by

the Initiative to Develop African Research Leaders (IDeAL) through the DELTAS

Africa Initiative [DEL-15-003], an independent funding scheme of the African

Academy of Sciences (AAS) ’s Alliance for Accelerating Excellence in Science in

Africa (AESA) and supported by the New Partnership for Africa’s Development

Planning and Coordinating Agency (NEPAD Agency) with funding from the

Wellcome Trust [107769/Z/10/Z] and the U.K. government. The funders had no

role in drafting nor the decision for submitting this manuscript.

Competing interests

There are no conflicts of interest declared by the authors.

Data Availability

Data used for this manuscript are available in Harvard Dataverse at

https://dataverse.harvard.edu/dataset.xhtml?persistentId=doi:10.7910/DVN/

L7S8TK. Access applications can be made through the Data Governance

Committee with details available on www.kemri-wellcome.org, or email to

cgmrc@kemri-wellcome.org.

301

Ethical approval

This study received approval from the Oxford Tropical research ethics

committee (OXTREC) from the University of Oxford (Ref: 525–17) and the

Kenyan Medical Research Institute (Ref: KEMRI/SERU/CGMR-C//086/3450).

302

References

1. World Health Organization, Antimicrobial resistance: global report on surveillance.

2014: World Health Organization.

2. Jonas, O.B., et al., Drug-resistant infections: a threat to our economic future. World

Bank Group, Washington, DC, 2017.

3. Holmes, A.H., et al., Understanding the mechanisms and drivers of antimicrobial

resistance. The Lancet, 2016. 387(10014): p. 176-187.

4. Okeke, I.N., Laboratory systems as an antibacterial resistance containment tool in

Africa. African journal of laboratory medicine, 2016. 5(3): p. 497-497.

5. World Health Organization, Step-by-step approach for development and

implementation of hospital and antibiotic policy and standard treatment guidelines.

2011, WHO Regional Office for South-East Asia.

6. Mendelson, M., et al., Maximising access to achieve appropriate human antimicrobial

use in low-income and middle-income countries. The Lancet, 2016. 387(10014): p. 188-

198.

7. Maina, M., et al., Antibiotic use in Kenyan public hospitals: Prevalence, appropriateness

and link to guideline availability. International Journal of Infectious Diseases, 2020. 99:

p. 10-18.

8. Livorsi, D., et al., Factors Influencing Antibiotic-Prescribing Decisions Among Inpatient

Physicians: A Qualitative Investigation. Infect Control Hosp Epidemiol, 2015. 36(9): p.

1065-72.

9. De Souza, V., et al., A qualitative study of factors influencing antimicrobial prescribing

by non-consultant hospital doctors. J Antimicrob Chemother, 2006. 58(4): p. 840-3.

10. Maina, M., et al., Inappropriate prescription of cough remedies among children

hospitalised with respiratory illness over the period 2002-2015 in Kenya. Trop Med Int

Health, 2017. 22(3): p. 363-369.

11. Ministry of Health. Basic Paediatric Protocols for ages up to 5 years. . 2016; Available

from: http://idoc-

africa.org/images/documents/2016/Basic_Paediatric_Protocol_2016/MAY%2023rd%

20BPP%202016%20SA.pdf.

12. Ramakrishnan, K., R.C. Salinas, and N.I.A. Higuita, Skin and soft tissue infections.

American family physician, 2015. 92(6): p. 474-483.

13. Organization, W.H., The selection and use of essential medicines: report of the WHO

Expert Committee, 2017 (including the 20th WHO Model List of Essential Medicines and

the 6th Model List of Essential Medicines for Children). 2017: World Health

Organization.

303

14. Ministry of Medical Services and Ministry of Public Health and Sanitation, Clinical

Management and Referral Guidelines – Volume III: Clinical Guidelines for Management

and Referral of Common Conditions at Levels 4–6: Hospitals. 2009, Ministry of Medical

Services and Ministry of Public Health and Sanitation,: Nairobi.

15. Kenyatta National Hospital and University of Nairobi, The KNH guide to empiric

antimicrobial therapy 2018: Nairobi.

16. Metlay, J.P., et al., Diagnosis and treatment of adults with community-acquired

pneumonia. An official clinical practice guideline of the American Thoracic Society and

Infectious Diseases Society of America. American journal of respiratory and critical care

medicine, 2019. 200(7): p. e45-e67.

17. Stevens, D.L., et al., Practice guidelines for the diagnosis and management of skin and

soft tissue infections: 2014 update by the Infectious Diseases Society of America. Clinical

infectious diseases, 2014. 59(2): p. e10-e52.

18. Penney, G.C., Adopting and adapting clinical guidelines for local use. The Obstetrician

& Gynaecologist, 2007. 9(1): p. 48-52.

19. Lo, B. and M. Field, Conflicts of interest and development of clinical practice guidelines.

Conflict of Interest in Medical Research, Education, and Practice, 2009.

20. Damiani, G., et al., The effectiveness of computerised clinical guidelines in the process

of care: a systematic review. BMC Health Services Research, 2010. 10(1): p. 2.

21. World Health Organization, WHO handbook for guideline development. Geneva: WHO;

2012. 2017.

22. Irimu, G., et al., Developing and introducing evidence based clinical practice guidelines

for serious illness in Kenya. Arch Dis Child, 2008. 93(9): p. 799-804.

23. English, M., et al., Developing guidelines in low-income and middle-income countries:

lessons from Kenya. Archives of Disease in Childhood, 2017. 102(9): p. 846-851.

24. Akech, S., et al., Magnitude and pattern of improvement in processes of care for

hospitalised children with diarrhoea and dehydration in Kenyan hospitals participating

in a clinical network. Tropical Medicine & International Health, 2019. 24(1): p. 73-80.

25. Gachau, S., et al., Prevalence, outcome and quality of care among children hospitalised

with severe acute malnutrition in Kenyan hospitals: A multi-site observational study.

PLOS ONE, 2018. 13(5): p. e0197607.

304

305

Chapter 8

Using a common data platform to facilitate

audit and feedback on the quality of hospital

care provided to sick newborns in Kenya

Michuki Maina, Jalemba Aluvaala, Paul Mwaniki, Olga Tosas-

Auguet, Catherine Mutinda, Beth Maina, Constance Schultsz

and Mike English BMJ Glob Health. 2018;3(5): e001027

306

ABSTRACT

Essential interventions to reduce neonatal deaths that can be effectively

delivered in hospitals have been identified. Improving information systems may

support routine monitoring of the delivery of these interventions and outcomes

at scale. We used cycles of audit and feedback (A&F) coupled with the use of a

standardised newborn admission record (NAR) form to explore the potential for

creating a common inpatient neonatal data platform and illustrate its potential

for monitoring prescribing accuracy. Revised NARs were introduced in a high

volume, neonatal unit in Kenya together with 13 A&F meetings over a period of

3 years from January 2014 to November 2016. Data were abstracted from

medical records for 15 months before introduction of the revised NAR and A&F,

and during the three years of A&F. We calculated, for each patient, the

percentage of documented items from among the total recommended for

documentation and trends calculated over time. Gentamicin prescribing

accuracy was also tracked over time. Records were examined for 827 and 7336

patients in the pre-A&F and post-A&F periods, respectively. Documentation

scores improved overall.

Documentation of gestational age improved from <15% in 2014 to >75% in

2016. For five recommended items, including temperature, documentation

remained <50%. 16.7% (n=1367; 95% CI 15.9 to 17.6) of the admitted babies

had a diagnosis of neonatal sepsis needing antibiotic treatment. In this group,

dosing accuracy of gentamicin improved over time for those under 2 kg from

60% (95%36.1 to 80.1) in 2013 to 83% (95% CI to 92.3) in 2016. We report that

it is possible to improve routine data collection in neonatal units using a

standardised neonatal record linked to relatively basic electronic data

collection tools and cycles of A&F. This can be useful in identifying potential

gaps in care and tracking outcomes with an aim of improving the quality of care.

307

Key Words: Neonatal Care, Audit and Feedback, Quality of care

Summary Box

• Improving information systems that support routine monitoring of

quality and outcomes at scale is an important part of efforts to

enhance neonatal care.

• We highlight clinical data elements that are poorly recorded by

practitioners in routine settings, findings that can help revise the

standardised record form.

• It is possible to improve routine data collection and prescribing

accuracy in neonatal units using a standardised neonatal record

linked to relatively basic electronic data collection tools and cycles of

audit and feedback to improve medical care.

308

INTRODUCTION

Newborn deaths account for approximately 44% of under-five deaths globally

[1] largely attributable to preterm birth, sepsis and intrapartum

complications.[2] Specific interventions such as newborn resuscitation, thermal

care, use of oxygen and early recognition and treatment of neonatal infections

have been identified as major interventions to reduce neonatal deaths that can

be effectively delivered as part of basic hospital services.[3] However, there are

few data on whether such interventions are delivered in routine settings in low

and middle-income countries (LMIC). Available evidence suggests that

adherence to recommended forms of care is poor.[4] For example looking at

treatment of neonatal infections, in a recent assessment of neonatal units in

Kenyan hospitals, more than 20% of the prescriptions of gentamicin were above

safe, recommended doses.[5]

An ability to monitor routine antibiotic prescribing is also aligned with

increasing concern over antimicrobial resistance (AMR). Neonatal units,

including intensive care units can be hotspots for development and

transmission of antimicrobial-resistant organisms. [6] This may be due to the

extensive empirical use of antimicrobial agents, inappropriate choice of

antibiotics, inappropriate dosing and extended duration of administration

coupled with poor infection prevention and control practices.[7-9] Spread of

antibiotic-resistant organisms amongst neonates may subsequently be

manifest in increased length of hospital stay, increased hospital costs and

greater morbidity and mortality.

Unfortunately, poor documentation, record keeping and information systems

[10] preclude effective monitoring of both the delivery of effective

interventions in general and the use of antibiotics in particular. This is worsened

by limited human resource capacity, equipment and supplies.[11] Improving

309

information systems that support routine monitoring of quality and outcomes

at scale is therefore, an important part of efforts to enhance neonatal care in

LMIC.[12]

To address the challenge of inadequate record keeping and routine monitoring,

we worked with a high volume, low-resource neonatal unit in Kenya using audit

and feedback (A&F). This was coupled with efforts to develop/update a

neonatal admission record (NAR) that could support a common inpatient

neonatal data platform for monitoring care and outcomes at scale. A&F is

based on the premise that if clinicians are informed of what is not consistent

with required practice, they will change behaviour. However, effects of audit

and feedback as an improvement strategy have been varied with some

reporting very modest or no effects. [13] A&F may be more effective if based

on data that are valid and timely [14] with recent and prior work suggesting

effectiveness is enhanced if coupled with other interventions such as the use of

standardised admission records. [15]

Here we report on the effects of repeated cycles of audit and feedback linked

to use of standardised record forms on the completeness of patient level

information in a LMIC neonatal unit. We illustrate the potential value of better

data for monitoring quality and prescribing accuracy and we identify which

clinical data elements are poorly recorded by practitioners in routine settings,

findings that can help revise the standardised record form.

Context: Pumwani Maternity Hospital

Pumwani Maternity Hospital, the largest public maternity hospital in Kenya, is

located in the capital city Nairobi. The hospital, which serves mainly the urban

poor population, conducts approximately 22, 000 deliveries annually; the

newborn unit has approximately 4500 admissions annually and holds close to

60 babies each day. Care is overseen by four consultant paediatricians

310

supervising a team of six medical officers and four clinical officers (non-

physician clinicians) typically working so that there are 2-3 clinicians in every

shift. Nursing care is provided by registered nursing officers a minority of whom

have specialised neonatal training. Typically, only 2-3 nurses are on duty per

shift assisted by trainee nurses. The research team had no role in patient care

but did support the provision of one clerk to collect data daily.

Description of routine data Collection using a newborn admission record

A NAR promoting documentation of key patient characteristics at the time of

admission was originally developed in 2006 as part of the Emergency Treatment

and Triage plus admission approach, which includes skill training in essential

inpatient newborn care.[16, 17] Adoption of this NAR has been at the discretion

of hospital teams with modest uptake but a suggestion that it can improve data

availability.[5] In Pumwani there was an effort to revise the NAR in 2014 with

the local team, so it might better capture important information, for example,

maternal human immunodeficiency virus (HIV) status, length of gestation and

mode of delivery. The NAR is divided into different sections which include: (1)

Relevant maternal history, (2) Babies’ biodata and clinical history, (3) Babies’

examination findings and admission vital signs, (4) the basic laboratory tests

ordered, and (5) primary and secondary diagnosis on admission. The clinical

variables included are based on the key signs and symptoms that national

guidelines recommend should be assessed for all sick newborns. [18, 19] A

minimum data set (used for the national reporting system) is collected on all

patients while a full data set including all clinical and treatment data is collected

on an automatically generated random sample of 60% of admitted newborns.

Table 1 shows the data collected from the NAR that were used for the analysis

in this report and some of the variables that were added after the modification

311

of the NAR in 2014. The complete NAR is attached in the online supplementary

appendix 1.

Table 1 NAR variables used for analysis.

Domain Number of

Variables

Variables included in the analysis

Maternal

History*

6 Mothers age, parity, gravidity, mothers blood

group, HIV status, VDRL status

Demographics

and diagnosis

12 Admission Information

Dates of birth, admission date, gender, birth

weight, age in days, gestational age, mode of

delivery, APGAR score at 5 minutes, admission

diagnosis.

Discharge information

Date of discharge or death, outcome (dead or

alive), diagnosis on discharge or death.

Presenting

complaints

6 Presence of fever, convulsions, difficulty in

breathing, vomiting, difficulty feeding, apnoea

Cardinal signs

on

examination

4 Grunting, central cyanosis, bulging fontanelle,

floppy (inability to suck, reduced

movements/activity)

Other physical

examination

18 Airway; stridor*

Breathing; bilateral air entry on chest

examination*, crackles on chest examination,

chest indrawing.

Circulation; skin pinch, femoral pulses*, capillary

refill time, heart murmur*, pallor, peripheral to

central skin temperature gradient *.

Others; signs of eye infection, signs of umbilical

infection, Presence of skin rashes/pustules, stiff

neck, irritability, jaundice, gestational size*,

severe wasting*,

Vital signs 4 Temperature, respiratory rate, heart rate,

oxygen saturation*

• Variables in modified NAR after 2014

312

Electronic data capture from the medical records which are in paper form (NAR,

treatment sheets,) occurs at discharge. Every working day data were collected

and managed using Research Electronic Data Capture (REDCap) electronic data

capture tools. REDCap® is a secure, web-based application designed to support

data capture for research studies.[20] These data are abstracted by a trained

clerk stationed at the health facility. To ensure data quality, built-in range and

validity checks are employed at the point of data entry, an automated error

checking procedure is run daily on site with corrections made. Further error

checks are performed by the data management team after de-identified data

are synchronised to a central server. A more comprehensive description of the

data cleaning and data quality assurance process is described elsewhere.[21]

Implementation activities

For a period of 15 months from January 2013 to March 2014, the period before

the A&F process, we collected baseline data in a retrospective survey. Data

were collected from records that included an earlier form of the NAR already in

use at the hospital. The records examined were from a random selection of

dates across the 15 months with the intention of capturing data on

approximately 55 admissions per month. From April 2014, the research team

worked with the neonatal unit clinical team to revise the NAR tailoring it to the

needs of the hospital a process resulting in the addition of further variables

(maternal history, addition of clinical signs and symptoms, addition of

admission diagnosis, table 1). The hospital team introduced the revised NAR

into its pre-printed medical files to make it a routine medical record filled for all

admissions. A&F was used to highlight documentation of key variables showing

how the NAR was used and the completeness of documentation. A&F was

313

integrated into existing monthly mortality meetings organised by the hospital

teams and attended by the clinical, nursing teams and hospital management

where these data were presented quarterly. Areas for improvement were

found and actions to promote change identified – with responsibility for leading

these left to the paediatricians and the clinical team.

Experience before and after A&F

The period for these data was divided into two phases, the baseline data

collection period January 2013 to 31 March 2014 and the period with feedback

from April 2014 to November 2016. We defined the 6-month periods, two pre-

A&F and six post A&F, to explore any effect of efforts to improve

documentation. For each patient, each variable within an analysis domain was

assigned a binary score denoting availability of documentation (i.e., 0= No; 1=

Yes). We then calculated, for each analysis domain, the percentage of

documented variables among the total number of variables that could be

documented for the patient population over the 6-month period. The

percentage of completed documentation was plotted with 95% (CI) to examine

changes and trends over time.

To understand what documentation tasks, remain difficult despite A&F, we

calculated, for each individual variable, the percentage of patients with missing

documentation for that variable over the last 6-month interval (i.e. April to

November 2016) and stratified items as poorly (25%-50%) or very poorly (<25%)

documented.

We provide documentation trends for individual variables that are

indispensable to delivering appropriate drug dosing to newborns-namely birth

weight and gestational age-and illustrate the impact of improved

documentation on trends of gentamicin posology over time to illustrate the

314

potential monitoring value of better data. The Kenyan guidelines recommend a

once daily gentamicin dose of 3 mg/kg for those babies <2 kg and 5 mg/kg for

those ≥ 2 kg in the first 7 days of life [19]. We considered a correct dose to be

within a ± 20% margin of error. Summary statistics and data visualisation were

conducted in R statistical software V. 1.0.136.

NARs were examined for 827 patients in the pre-A&F period (January 2013–

March 2014). The revised NAR was examined for 7336 patients in the A&F

period, over six consecutive 6-month intervals (n=1067 (Period 1, 2014),

n=1941 (Period 2, 2014), n=1144 (Period 1,2015), n=1103 (Period 2, 2015),

n=1230 (Period 1, 2016) and n=680 (Period 2, 2016)). The last 6-month period

in 2016 had lower patient numbers due to a 14-day doctors strike in the month

of October. Of the 7985 patients included, 46% (n=3656) were female, 78.3%

(n=6251;95% CI 77.6 to 79.4) were admitted on the date of birth, 90.5%

(n=7207; 95% CI 89.8 to 91.1) were admitted within 48 hours of delivery and

9% (n=712; 95% CI 8.3 to 9.6) of the admitted babies died. The mean birth

weight on admission was 2.88 kg (95% CI 2.86 to 2.90). There were 13 A&F

feedback meetings between January 2014 and November 2016, eight specific

to A&F feedback and five as part of monthly morbidity and mortality meetings.

Documentation across domains

The documentation across all domains showed improvement over time and

after introduction of the revised NAR, as shown in figure 1. The greatest

improvement was noted in the domains that had low baseline performance at

the onset. Thus, documentation of babies’ vital signs, maternal history and

other physical examination showed greatest improvement. Although

documentation of vital signs improved by more than 50% between 2014 and

2016 performance stagnated at less than 75%, as it did in another domain

‘other physical signs’ (table1).

315

In the variable-specific analysis over the last 6-month period (April-November

2016), we explored documentation of 57 variables in 680 patients. Variables

with poor (25%-50%) documentation were general examination of the skin

(other clinical signs) and admission temperature (vital signs). Those with very

poor documentation (<25%) were all under other physical examination; these

were; documentation of visible wasting (a possible indicator of intrauterine

growth retardation), skin pinch and skin temperature (signs of neonatal

dehydration,) and the presence of femoral pulses and eye discharge. This

variable specific analysis helps explain the plateauing seen in figure 1.

Figure 1 Trends in completeness of documentation of various parts of the neonatal admission

record, (NAR) over three years. The black shade around the trends is the 95% confidence intervals

around the estimates. The dotted line represents the introduction of prospective data collection

in April 2014.

316

Documentation of gestational age and birth weight

Documentation of birth weight and gestational age are important in feed and

antibiotic prescribing decisions and other aspects of newborn care. The

documentation of birth weight remained consistently high with >95%

documentation since 2013. Documentation of gestational age has shown

gradual improvement over the data collection period from < 15%

documentation in 2014 to >75% documentation in 2016, as shown in figure 2

below.

Figure 2 Shows the proportion of documentation of gestational age among hospitalised neonates

in six-month intervals with 95% confidence intervals around the estimates. The dotted line

represents the introduction of prospective data collection.

Dosing of gentamicin

There were 16.7% (n=1367; 95% CI 15.9 to 17.6) of the admitted babies with a

diagnosis of neonatal sepsis needing antibiotic treatment during the study

317

period. Of these neonates diagnosed with neonatal sepsis, 81.4% (n=1113; 95%

CI 79.2 to 83) were prescribed a combination of penicillin and gentamicin

(national first line). An additional 39.5% (n=2688; 95% CI 38.4 to 40.7) of

admitted neonates were put on antibiotic treatment without a clear diagnosis

of neonatal sepsis who typically had respiratory distress, prematurity or

asphyxia; these prescriptions were predominantly penicillin and gentamicin. Of

the neonates who received gentamicin, 4.4% (n=161;95% CI 3.8 to 5.1) received

an overdose and 2.4% (n=86;95% CI 1.9 to 2.9) were underdosed. By plotting

trends by weight, the major changes were noted in those <2 kg with a reduction

in the proportion of overdoses of gentamicin prescribed, as shown in figure 3

below.

Figure 3: Charts showing dosing trends of gentamicin over time in six-month intervals from

January 2013 for those under 2 kg (Upper panel) or over 2 kg (lower panel). The grey shade

indicates the 95% confidence intervals around the estimates. The dotted line represents the

introduction of prospective data collection.

318

Impact of introducing the NAR with A&F cycles

We set out to improve availability of clinical data on neonatal admissions in a

large Kenyan hospital. Here we report how repeated cycles of A&F were used

to improve availability of information, share lessons on information that is

difficult to collect in busy routine clinical settings and illustrate how such data

may be of potential value for monitoring quality of care, taking the example of

gentamicin dosing. We focused on reinforcing the use of a standardised

admission record as these have been associated with more thorough

documentation in some LMIC settings but not previously, to our knowledge, in

neonatal care.[22] We report good documentation in six domains that include

50 specific variables. However, five specific variables were typically poorly

documented by clinicians; examination of the skin (colour, bruising, pustules),

admission temperature, visible wasting, skin pinch and skin temperature.

Clinicians may feel some of these are less relevant to the babies’ clinical

condition on the day of birth (wasting, skin pinch and skin temperature) when

most admissions occur as these signs are typically associated with later onset

neonatal illnesses. With thermal care being an essential aspect of neonatal

care, we note that temperature is still recorded poorly with approximately 25%-

50% documentation. Documentation of gestation was noted to improve across

the period to >75% documentation. This is much better than previous reports

from Kenyan hospitals.[5] Here we also report some improvement in the

accuracy of gentamicin dosing for neonates under 2 kg. Dosing errors, including

antibiotics occur more among sick neonates compared to any other population.

These errors may have a more significant effect as neonates have little

physiological ability to buffer these errors. [23] Most of these errors occur in

the prescribing phase as compared with the dispensing and administration

phase. Internationally, various interventions have been instituted to reduce

319

these prescriptions errors. [24-26] Our data demonstrate it is possible to track

correctness of dosing over prolonged periods in a busy hospital setting.

A&F as an improvement tool has been widely used in clinical settings, but

evidence on its effects is mixed [13]. It may be more effective if the

performance targeted has large room for improvement. This was sometimes

the case in our setting where, for example, documentation of gestation was <

20% in the baseline period of our study. It has also been noted that A&F is more

effective when the targeted change is less complex (requiring no specific skills)

and compatible with clinician norms and values.[27] Having the organisational

buy in and involving the leadership in goal and target setting also makes the

process of A&F more effective.[28] The goal setting and feedback meetings in

our setting were attended by the heads of the units and hospital managers,

which ensured that the process was in line with overall hospital priorities.

Paucity of newborn data in terms of type of care provided, morbidity and

mortality is still a major challenge in many facilities.[29] The WHO names

actionable information systems as a key pillar in provision of quality maternal

and neonatal care. This allows clinicians to make timely and appropriate

decisions.[12] There is increasing interest at the global level in tracking quality

of care at scale, particularly care provided to the newborn[30]. A central

element of monitoring quality of care at scale is a common data set. Our work

demonstrates that clinicians providing care in a busy, routine hospital setting

can be encouraged to use standardised neonatal record forms with high levels

of documentation, especially for variables that are clinically meaningful to

them. Establishing an agreed and standard medical record could enable

neonatal networks to be formed with an aim of improving care provided to

patients. In such networks, colleagues may also share experiences and can

provide a ‘bottom up’ method of problem solving [31] helping improve clinical

320

outcomes at scale [32]. With the realisation that quality data may improve care

provision, many governments and hospitals in Africa are moving towards

electronic medical records (EMR).[33] This may provide an opportunity to

integrate agreed common data elements as part of an EMR as well as improve

the value of existing District Health Information systems.[34]

Limitations of this approach are that documentation only captures some

aspects of quality and that information can be documented incorrectly or

activities recorded but not done. However, data from case records remain the

most feasible form of information to collect on patients at scale. The accuracy

of data is also hard to verify for certain types of data. For example, most of the

documentation of gestation is still based only on maternal history rather than

ultrasound dating in Kenya. Other challenges to implementing better

information systems remain; the resources needed to support data capture and

analysis with better information are often given low priority in resource

allocation. While data collection for this project was partly supported by the

research team, this was limited to the A&F reports and the modest costs for

data collection (clerk costs). However, there are discussions with the ministry

of health and the national paediatric association on how best to transition and

expand these activities to ensure continuity beyond the current funding.

Despite this being a research project, outputs from this work feed into the

routine mortality and morbidity meetings and monthly health information

reports further reinforcing data use and institutionalisation of the activities

beyond the current funding cycle.

321

CONCLUSION

It is possible to improve routine data collection in neonatal units using a

standardised neonatal record linked to relatively basic electronic data

collection tools. These tools could collect data on significant presenting

complaints, maternal history, babies’ physical examination, investigations, and

treatment. Data collected on such a platform at wider scale can be useful in

identifying potential gaps in care with an aim of improving the quality of care

provided in facilities and tracking outcomes. Monitoring antibiotic use could be

especially valuable in the current era. Implementing such systems takes time

and needs significant support from clinicians, nurses and hospital managers.

Authors’ contributions

The roles of the contributors were as follows: M.M, J.A and M.E conceived the

study. P.M, OT assisted M.M in analysis and interpretation of these data. M.M,

J.A, O.T, C.S and M.E drafted the manuscript. J.A, O.T, C.S and M.E critically

revised the manuscript for intellectual content. All authors read and approved

the final manuscript.

Acknowledgements

The authors would like to thank the Pumwani Maternity Hospital who gave

permission for the study included in this paper. We also thank the hospital

paediatricians and clinical teams on newborn unit who provide care to the

neonates for whom this project is designed. This work is published with the

permission of the director of KEMRI.

Funding

This work was supported by funds from The Wellcome Trust (#097170)

awarded to ME to support M.M, P.M, J.A and O.T. M.M was also supported

through the DELTAS Africa Initiative [grant# DEL-15-003]. The DELTAS Africa

322

Initiative is an independent funding scheme of the African Academy of Sciences

(AAS)’s Alliance for Accelerating Excellence in Science in Africa (AESA) and

supported by the New Partnership for Africa’s Development Planning and

Coordinating Agency (NEPAD Agency) with funding from the Wellcome Trust

[grant # 107769/Z/10/Z] and the UK government. The funders had no role in

drafting or submitting this manuscript.

Competing interests

There are no conflicts of interest declared by the authors.

Ethical approval

The findings reported here come from a study that was approved by the KEMRI

Scientific and ethics review committee (Approval No SERU/3459).

323

References

1. UN Inter-agency Group for Child Mortality Estimation, Levels and trends in child

mortality 2014. 2014.

2. Black, R.E., et al., Global, regional, and national causes of child mortality in 2008: a

systematic analysis. Lancet, 2010. 375(9730): p. 1969-87.

3. Bhutta, Z.A., et al., Can available interventions end preventable deaths in mothers,

newborn babies, and stillbirths, and at what cost? Lancet, 2014. 384(9940): p. 347-70.

4. Wilunda, C., et al., Assessing Coverage, Equity and Quality Gaps in Maternal and

Neonatal Care in Sub-Saharan Africa: An Integrated Approach. PLOS ONE, 2015. 10(5):

p. e0127827.

5. Aluvaala, J., et al., Assessment of neonatal care in clinical training facilities in Kenya.

Archives of Disease in Childhood, 2015. 100(1): p. 42-47.

6. Russell, A.B., M. Sharland, and P.T. Heath, Improving antibiotic prescribing in neonatal

units: time to act. Archives of Disease in Childhood - Fetal and Neonatal Edition, 2012.

97(2): p. F141-F146.

7. Hsu, J.F., et al., Predictors of clinical and microbiological treatment failure in neonatal

bloodstream infections. Clin Microbiol Infect, 2015. 21(5): p. 482.e9-17.

8. Patel, S.J. and L. Saiman, Principles and strategies of antimicrobial stewardship in the

neonatal intensive care unit. Semin Perinatol, 2012. 36(6): p. 431-6.

9. Goldmann, D.A., et al., Strategies to Prevent and Control the Emergence and Spread of

Antimicrobial-Resistant Microorganisms in Hospitals. A challenge to hospital

leadership. Jama, 1996. 275(3): p. 234-40.

10. Aluvaala, J., et al., Delivery outcomes and patterns of morbidity and mortality for

neonatal admissions in five Kenyan hospitals. Journal of Tropical Pediatrics, 2015. 61(4):

p. 255-259.

11. English, M., et al., An intervention to improve paediatric and newborn care in Kenyan

district hospitals: understanding the context. Implement Sci, 2009. 4: p. 42.

12. World Health Organization, Standards for improving quality of maternal and newborn

care in health facilities. 2016, WHO: Geneva. p. 19-21.

13. Ivers, N., et al., Audit and feedback: effects on professional practice and healthcare

outcomes. Cochrane Database of Systematic Reviews, 2012(6).

14. Ivers, N.M., et al., No more ‘business as usual’ with audit and feedback interventions:

towards an agenda for a reinvigorated intervention. Implementation Science, 2014.

9(1): p. 14.

324

15. Gachau, S., et al., Does audit and feedback improve the adoption of recommended

practices? Evidence from a longitudinal observational study of an emerging clinical

network in Kenya. BMJ Global Health, 2017. 2(4).

16. Irimu, G., et al., Developing and introducing evidence based clinical practice guidelines

for serious illness in Kenya. Arch Dis Child, 2008. 93.

17. English, M., et al., Implementing locally appropriate guidelines and training to improve

care of serious illness in Kenyan hospitals: a story of scaling-up (and down and left and

right). Arch Dis Child, 2011. 96(3): p. 285-90.

18. World Health Organization, Pocket Book of Hospital Care for Children: Guidelines for

the Management of Common Childhood Illnesses. 2013, World Health Organization

Geneva.

19. Ministry of Health. Basic Paediatric Protocols for ages up to 5 years. 2016; 4:[Available

from: http://idoc-

africa.org/images/documents/2016/Basic_Paediatric_Protocol_2016/MAY%2023rd%

20BPP%202016%20SA.pdf.

20. Harris, P.A., et al., Research electronic data capture (REDCap)--a metadata-driven

methodology and workflow process for providing translational research informatics

support. J Biomed Inform, 2009. 42(2): p. 377-81.

21. Tuti, T., et al., Improving documentation of clinical care within a clinical information

network: an essential initial step in efforts to understand and improve care in Kenyan

hospitals. BMJ Glob Health, 2016. 1(1): p. e000028.

22. Mwakyusa, S., et al., Implementation of a structured paediatric admission record for

district hospitals in Kenya--results of a pilot study. BMC Int Health Hum Rights, 2006. 6:

p. 9.

23. Kaushal, R., et al., Medication errors and adverse drug events in pediatric inpatients.

Jama, 2001. 285(16): p. 2114-20.

24. Radley, D.C., et al., Reduction in medication errors in hospitals due to adoption of

computerized provider order entry systems. J Am Med Inform Assoc, 2013. 20(3): p.

470-6.

25. Leape, L.L., et al., Pharmacist participation on physician rounds and adverse drug events

in the intensive care unit. Jama, 1999. 282(3): p. 267-70.

26. Manias, E., A. Williams, and D. Liew, Interventions to reduce medication errors in adult

intensive care: a systematic review. Br J Clin Pharmacol, 2012. 74(3): p. 411-23.

27. Foy, R., et al., Attributes of clinical recommendations that influence change in practice

following audit and feedback. Journal of Clinical Epidemiology. 55(7): p. 717-722.

325

28. Ivers, N., et al., Feedback GAP: study protocol for a cluster-randomized trial of goal

setting and action plans to increase the effectiveness of audit and feedback

interventions in primary care. Implement Sci, 2013. 8.

29. Gathara, D., et al., Quality of hospital care for sick newborns and severely malnourished

children in Kenya: a two-year descriptive study in 8 hospitals. BMC Health Serv Res,

2011. 11: p. 307.

30. WHO. UNICEF, Every Newborn: an action plan to end preventable deaths, W.H.

Organization, Editor. 2014: Geneva.

31. Haines, M., et al., Determinants of successful clinical networks: the conceptual

framework and study protocol. Implementation Science, 2012. 7(1): p. 16.

32. Vermont Oxford Network. What is Vermont Oxford Network? [cited 2017 25th January

2017]; Available from: https://public.vtoxford.org/about-us/.

33. Akanbi, M.O., et al., Use of Electronic Health Records in sub-Saharan Africa: Progress

and challenges. J Med Trop, 2012. 14(1): p. 1-6.

34. Ministry of Health. Kenya Health information system. [cited 2018 10th April ]; Available

from: https://hiskenya.org/.

326

Supplementary Materials Neonatal Admission Record

327

328

329

Chapter 9

Discussion

330

DISCUSSION

This thesis describes the infection prevention and control and antibiotic

stewardship capacities in Kenyan public hospitals. I sought to explore the

current status using surveys and additionally used qualitative approaches to

understand the context better. Though the findings are specific to Kenya, the

lessons learned herein apply to other low and middle-income countries with

similar health systems and facing similar challenges with IPC and ABS [1]. In

addition, this work generated a WASH Survey tool that can be used in other

countries to assess IPC and WASH.

Why this work is important

Whereas there have been some assessments of IPC carried out in Kenya in

some selected smaller health facilities [2], this work includes a larger sample of

hospitals with bed capacities ranging between 130 and 600 beds (Chapter 2

and 3 ). It, therefore, paints a broader picture of the status in Kenya as these

hospitals are also spread across the country, covering areas with differences in

the burden of various infectious diseases. The WASH-FAST, which was modified

from the original WASH facility assessment tool (WASH-FIT) offers a more

comprehensive approach to assessing WASH in large facilities since it provides

for assessment at both ward and facility levels (Chapter 2). The WASH-FAST

also offers an opportunity for audit and feedback. Through repeat assessments

in a facility, the IPC teams can identify what needs urgent intervention and

make recommendations to the hospital managers. This is particularly important

in settings with ongoing or emerging disease outbreaks such as COVID-19. In

such outbreaks, application of these tools can help identify in a systematic and

timely manner any shortcomings or breakdowns in IPC and/or WASH

infrastructure or practice, including poor hand hygiene facilities.

331

The original WASH surveys were carried out in 2018. After feedback to the

hospitals, some undertook improvements which have proved to be timely with

the emergence of the COVID-19 pandemic in 2020. The pivotal role played by

IPC including hand hygiene and the use of personal protective equipment in the

management of COVID-19 has been demonstrated by the numerous cases of

health care workers contracting COVID-19 in the line of duty globally [3]. The

pandemic has, therefore provided an opportunity to strengthen IPC structures

and leadership to deal with COVID-19 effectively. This was illustrated in the

follow-up phone interviews at the onset on the COVID -19 pandemic in Kenya.

From these interviews, some hospitals were better prepared to deal with

COVID-19 since they took action to improve WASH after the initial surveys

(Chapter 4). Nevertheless, these investments should be sustained beyond the

pandemic as IPC improves patient and health worker safety for all other

infectious diseases.

In settings like Kenya, where resources to run and improve hospitals are limited,

the WASH surveys allow for better decision making as these funds can be

directed to the most critical interventions. Additionally, with a comprehensive

understanding of the status of WASH in a hospital, the managers can justify

funding proposals to the county governments or development partners.

Improving accountability for most health system components remains a

challenge in many settings. The WASH-FAST addressed this by identifying those

responsible for making the necessary improvements making follow up possible

and enhancing accountability (chapter 2). To ensure these responsibilities are

taken seriously by the relevant staff, they should be included in the staff job

roles and performance appraisals. As part of improving accountability, the roles

played by the IPC committees and focal persons have also been highlighted in

this work (chapter 3). These roles have been vital during the COVID-19

332

pandemic for the provision of leadership and technical support to hospital

management (chapter 4). These teams and committees need to be formally

appointed, trained and well-resourced in all facilities to ensure proper IPC

management.

This work has also highlighted the gaps in leadership, training and funding for

antibiotic stewardship activities (chapter 5). The lack of up to date treatment

guidelines for clinicians to use, especially in the adult and surgical units, has also

been noted (chapter 6). The absence of guidelines was also reflected by the

different antibiotics used to treat similar conditions and the high proportion of

inappropriate treatments accorded to the hospitalised patients. These

inappropriate treatments in a background of limited laboratory support and

weak ABS structures have the potential to fuel antimicrobial resistance

significantly [4].

Additionally, I make a case for the development of clinical practice guidelines in

the units where these are absent or outdated. Priority should be for the most

common conditions that do not have guidelines, or the existing guidelines are

not up to date. The process of developing these guidelines should involve all

key stakeholders, including the Ministry of Health, professional bodies and

where possible patient groups. With lessons from the development of the basic

paediatric protocols in Kenya, any new guidelines should also include a training

and dissemination strategy (chapter 7).

The work also provides a perspective on the successful use of a relatively

inexpensive data capture platform. This platform collects data from routine

admission records. These data are useful for hospital audits and decision

making. This was shown to improve monitoring of antibiotic use, dosing

accuracy and clinical care in a busy neonatal unit (Chapter 8).

333

Key Insights

Funding for IPC activities

This work highlights persistent challenges in the funding of IPC activities in

public hospitals in Kenya. In some of the hospitals, there is no budget set out

for IPC activities, and these hospitals rely on funds from other hospital

departments to fund IPC activities. Even in the facilities where budgets for IPC

were prepared, minimal funds are actually allocated for these activities. This is

because the process of budget approval and allocation is done at the regional

government level. To put this into the broader Kenyan context, the devolution

of health services to the regional governments has not only reduced direct

funding to hospitals but also reduced the ability of hospitals to make the

necessary procurements [5]. Before devolution, the facilities were able to

operate hospital bank accounts from funds generated from the hospitals. These

funds were under the hospital management boards, and the hospitals were

allowed to incur expenditure and make necessary procurement. The process

was in some locations responsive to local needs and efficient, and hospitals

were able to fund activities like IPC, but this process was also riddled with

corruption and misappropriation not only in Kenya but in many countries that

use this model [6, 7]. The new constitution under the public finance

management Act in a bid to improve accountability directed that all funds at

county level be under the county finance teams[8]. This new financial

management system has reduced hospital autonomy and results in

procurement delays, and the county governments have the final say on what

can be procured. These checks and balances seek to reduce corruption and

misappropriation, but they result in unnecessary delays which affect service

delivery [5].

334

Low status and staff interest for IPC and related activities

The low status of IPC and related activities have also been highlighted. In the

survey hospitals, the staff members did not take an interest in learning about

or improving IPC. This was evidenced by the low turnout for IPC training

activities and the reluctance of senior clinicians to join the IPC committees. As

part of the WASH/IPC assessments (Chapter 3), we assessed for any rewards

systems or incentives for well-performing staff. These were absent in most of

the study hospitals. The lack of incentives and a reward system coupled with

poor staff remuneration and often high workloads may contribute to burnout

and be a significant contributor to this low morale and the perceived lack of

interest. Again, to put this in the context of the Kenyan health system, these

factors have also been reflected by the frequent health worker strikes

demanding better pay and working conditions. In 2017 there was a 100-day

doctors strike and 150-day nurses strikes[9]. Since then, there have been a

series of health worker strikes at the county level. To improve staff attitudes

and morale towards IPC and overall productivity, improving remuneration and

working conditions is crucial, indeed improving IPC is part of a key issue in

existing working conditions [10].

The ambiguity of IPC roles and responsibilities for health workers

As presented in chapter 3, in the hospital surveys, it was noted that staff did

not have IPC related activities in their job descriptions and appraisals. IPC roles

and responsibilities can be improved by including IPC related tasks on the staff

appraisal not only for the hospital managers but also for the other hospital staff.

From the available literature, the ambiguity of roles on IPC has been noted to

be a reason why some activities are not performed, and in many hospitals, has

resulted in hospital-acquired infections. With this ambiguity, the HAI's which

335

are primarily due to poor IPC are therefore perceived as unavoidable or

becomes a blame game with no one taking responsibility for their occurrence

[11] [12]. This is a situation that affects patient safety and best described by

Woods and colleagues as "the problem of many hands" [13]. This is where

multiple people or groups contribute to the success or failure of a situation, e.g.

IPC, but no one person can be held responsible. To mitigate this problem of

many hands in the setting of IPC, all hospital staff should have clear roles and

responsibilities as IPC concerns everyone, from the managers to the frontline

health workers.

As the roles and responsibilities for IPC are put in place in hospitals, there is a

need to be cognisant of how hierarchy across hospital staff would affect the

success of such an intervention. In hospitals with consultants, doctors, nurses

and other clinicians, it has been noted that the nurses are keener on the

practice of IPC. In chapter 3 of our work, this was confirmed by a nurse manager

reporting that nurses in her facility were more conversant with IPC than other

cadres. Although my work did not explore the reasons behind this, to explain

why this might be the case, previous work by Gilbert and colleagues seeking to

understand the ethics and politics of IPC, albeit in a high-income setting, noted

that senior clinicians perceive that some of the IPC policies are 'imposed' on

them and are a threat to their professional independence, power and

autonomy [11]. Since the IPC policies are meant to improve patient safety and

care, any clinician not keeping to these standards should be viewed as

unprofessional and not keeping to the ethics of doing no harm [11]. The

hospitals should provide mechanisms for all staff, including the juniors, to raise

their concerns about unprofessional behaviour, including inadequate infection

prevention and control practices by staff. Creating a team environment where

336

the seniors can learn and are also accountable to others can improve IPC

behaviour and status.

Hospitals outsourcing IPC services

In some of the hospitals we studied, the management had resorted to

contracting private companies to offer cleaning and waste disposal services as

described in chapter 3. From the available literature, this process of outsourcing

cleaning services is seen as not only a potentially cost-effective measure but

also a way to ensure that the hospital can focus on its core mandate of providing

care [14]. In the facilities where this was taking place, it was the prerogative of

the service provider to provide cleaning materials and staff and ensure the

facilities are clean at all times. However, these services need to be overseen by

hospital managers to ensure the cleaning companies provide excellent quality

services. The practice of outsourcing in hospitals has not always been

successful. Globally, in hospitals where outsourcing has failed, this has been

due to poor selection of providers and incomplete contracts that lead to poor

relations between the providers and the hospitals [14]. The hospitals,

therefore, need to select credible and experienced service providers and have

a clear contract on the roles and responsibilities. Hospitals cannot, however,

entirely abdicate the IPC responsibilities to contracted service providers.

Poor Antibiotic Stewardship leadership

A critical first step for hospitals in improving ABS in Kenyan hospitals, as

envisioned by the national ABS guidelines recently launched, is for hospitals to

constitute committees to provide leadership for ABS [15]. Just like in IPC, on

antibiotic stewardship, the gaps in leadership were also highlighted. From the

stewardship assessments (chapter 5), the hospitals lacked ABS committees. The

337

committee that was present in some of the facilities was the medicines and

therapeutics committees which was composed of clinicians, pharmacists and

hospital managers. The main aim of these therapeutic committees was to

provide direction on what medicines and technologies the hospital should

procure and did not focus on the stewardship of these treatments, including

antibiotics. With proper training, these therapeutics committees are a good

starting point for stewardship activities. Properly constituted stewardship

committees can ensure the clinicians are well trained on antibiotic stewardship

and prescribing. This can be done through routine continuous medical

education and staff orientation programmes. These committees are also able

to lobby for funding to carry out audit activities.

Pharmacists at the centre of stewardship activities

From interviews with the pharmacists in the participating hospitals, we noted

that most of these pharmacists were not actively involved in the hospital

stewardship activities and only a handful were involved in the clinical hospital

rounds and actively involved in patient care by advising the clinicians on the

antibiotic prescription choices, combinations and drug interactions (chapter 5).

To improve stewardship, the roles played by the pharmacists in public hospitals

need to go beyond drug dispensing. In Kenyan public hospitals, the bulk of the

prescriptions are done by the junior clinicians, some of whom are in training.

Involving the pharmacists in routine patient care, including training the junior

clinicians reduces dosing errors and improves prescription practices [16, 17].

In addition, we noted that the pharmacists in the survey hospitals have

opportunities to provide continuing medical education to the hospitals. These

opportunities can be used to promote antibiotic stewardship and rational

prescribing to all the clinicians.

338

Challenges with antibiotic supplies and quality

On drug supplies, there were noted frequent drug stockouts in the hospitals

(chapter 5). In some of the hospitals, the pharmacists blamed these stockouts

on the suppliers. From related literature, in Kenyan public hospitals, these

stockouts have been attributed to weak procurement structures where these

hospitals rely on the Kenya Medical Supplies Agency, which is the government

body mandated to supply drugs to the public hospitals. With these

arrangements, drugs are supplied at specific times of the year which often leads

to stockouts [18]. Poor forecasting and quantification of the hospital drug needs

also contribute to these stockouts [18].

From our surveys, as discussed in chapter 5, some of the clinicians were also

concerned about the quality of antibiotics available, especially those from the

private drug stores. In cases of antibiotic stockouts, where alternatives are not

available in the hospital, the patients have to source these drugs from private

pharmacies. Due to cost implications, these patients opt for the cheaper

options, some of which are poor quality antibiotics. From the broader

literature, the use of poor quality or falsified drugs, including antibiotics, may

lead to poor treatment outcomes and treatment failure [19].

Study Limitations

I present some limitations to this work. In the IPC surveys (chapter 3), all the 65

WASH indicators were provided with the same weight. Indeed, some indicators

are more critical to improving IPC, e.g. ensuring water is available in the

hospitals. Assigning different weights to these indicators would allow the

managers to prioritise the improvement of the critical indicators.

The original WASH-FIT and WASH-FAST offer a risk assessment activity as part

of the assessment. In this activity, the people conducting the survey are

provided with an opportunity to gauge through consensus the most critical

339

indicators that need urgent improvement. This risk assessment process is

tedious and therefore providing weights to the indicators beforehand might

make this process easier. Future work to improve the WASH FAST would include

weighting the indicators. A stakeholder engagement process similar to the one

previously carried out in our surveys can provide an excellent platform to assign

weights through consensus.

During the surveys , we excluded some wards that were not comparable across

the hospitals. These wards include critical care units where issues of IPC are

essential. For future surveys at the hospital level, including these wards is vital

for the government and hospital managers to have a good understanding of the

state of IPC in the facility. It also allows for accurate resource allocation to

improve IPC.

This work is based on a sample of 16 Kenyan public hospitals. Although these

hospitals provided a diverse range of public hospitals based on hospital size,

malaria and HIV prevalence, I acknowledge that these results may not be

representative of the IPC picture in private and faith-based hospitals. Future

work may consider including a larger sample of hospitals but also include

private and faith-based hospitals to have a more representative picture of the

exact state of IPC and ABS in Kenya

Behaviour is a crucial component to the success or failure of any intervention.

The surveys in chapter 3 to 6 did not assess or observe clinician behaviour which

is a pivotal contributor to the success of both IPC and ABS. The focus was more

on the infrastructure and organisational arrangements. Although the

interviews gave some insight into understanding behaviour, the observation of

practice would have complimented these findings. Future work would need to

factor in some observations of practices like hand washing, proper use of PPE

and waste segregation by the clinicians.

340

The use of a point prevalence approach (chapter 6 ) in assessing antibiotic use

also presented some limitations. Key to this is that the method does not allow

for the assessment of patient outcomes, which would be crucial in

understanding the effects of the inappropriate prescriptions. A longitudinal

survey would allow for the assessment of antibiotic consumption using the

recommended defined daily dose[20]. This approach, where time and

resources allow, provides a more comprehensive understanding of antibiotic

use and the costs of treatment. We focussed on antibiotic use in the inpatient

setting. I acknowledge that also understanding how the antibiotics are

prescribed in the ambulatory care settings and clinics in addition to the

inpatient units would provide a more holistic picture to help us understand

antibiotic use and stewardship. Review of outpatient antibiotic prescriptions

would also be a useful addition to the Global Point Prevalence Survey of

Antimicrobial Consumption and Resistance ( GLOBAL- PPS) approach that was

adapted for this work [21].

This work was focused on clinicians and hospitals and did not assess the patient

perceptions on IPC antibiotic use. Indeed as Sutton and colleagues correctly

state, IPC should be everybody's business and should encourage patient

involvement [22]. To ensure interventions on IPC and antibiotic use are

effective, a clear understanding of how patients perceive IPC, antibiotic use and

AMR is critical.

Utility and Next Steps

Policy generation and training

This work provides vital information for the policymakers in Kenya and other

settings with similar health systems. The Kenya national action plan for the

prevention and containment of AMR recommends the generation of local data

341

to track IPC and antibiotic use [23]. The WASH-FAST provides an excellent

starting point for hospitals to conduct IPC surveys for the audit of IPC structures.

These audits coupled with proper accountability as provided in the WASH-FAST

would potentially bring the desired improvements within the hospitals, and if

done at scale, these data are crucial for decision making and policy generation

at a national level.

These surveys involved collaboration with the Ministry of Health, and the

findings have been shared at national and regional government levels to assist

in policy generation work. Chapter five addendum provides a policy brief

prepared from the WASH surveys. For better policy generation and decision

making by the Ministry of Health, there is a need to scale up these surveys

across the country, including the private health facilities.

Our work highlights the need to further engage in training the health workers

to ensure all are conversant with IPC and antibiotic stewardship. I suggest the

design of an IPC/ABS training module for health workers which is accredited by

the Ministry of Health. This can be piloted and scaled up and become a

prerequisite for all health workers to receiving their practice licenses or

renewals. Such a course would ensure that all health workers are familiar with

the best practice. Such modules have been previously rolled out nationally

through open, and distance learning approaches to deal with HIV and more

recently with COVID-19 [24, 25]. Once the health workers are trained, the

hospital managers and county governments can include relevant IPC and ABS

roles and responsibilities on the staff job descriptions, which is currently lacking

in most hospitals.

342

Funding for IPC and ABS

In many low- and middle-income countries where resources are scarce, there

needs to be a clear intention to allocate funds towards IPC and ABS. For most

health managers, these interventions are not income-generating, and hence

the available funds are channelled to areas that will see a return on investment.

Whereas these interventions generate no direct income, the economic gains to

the patient and the health system, in general, are numerous. These include the

costs, time and effort related to treating hospital-acquired infections and the

social and economic effects of prolonged hospitalisation on the patient [26]. As

the funders of health care, the government through the Ministry of Health or

county governments and national health insurers have a role to play to ensure

laws and compel hospitals to earmark funds for IPC and ABS activities.

Improving IPC structures for hospital accreditation

Proper IPC structures need to be in place as part of the hospital accreditation

systems. In Kenya, adherence to key IPC policies is one of the criteria for the

accreditation of medical schools and teaching hospitals [27]. However, this

should be extended to all hospitals in Kenya. This process of IPC inspection can

be done regularly by an independent government body. The results of these

inspections can then be used for accreditation of hospitals by insurance

providers and also for licencing by the regulatory agencies [28]. In addition, to

improve accountability, including IPC roles and responsibilities on the staff, job

descriptions can be included as a criterion for hospital accreditation.

IPC and ABS guideline development and dissemination

Treatment guideline development is an area that needs more focus and input

from all stakeholders. From this work, we noted that outside the basic

paediatric protocols, there are few up to date guidelines for clinicians to rely on

343

in practice. More efforts need to focus on this process as a critical step in

improving antibiotic use and reducing AMR.

Looking at IPC, though there are national IPC guidelines in Kenya, we did not

find any efforts to provide training and dissemination to the hospitals. We

recommend wider dissemination and operationalisation to ensure they achieve

the desired effects.

Encouraging the use of data platforms and electronic medical records

Encouraging hospitals to adopt the use of electronic records can be a crucial

step to improve audit and feedback on drug utilisation, including antibiotics.

These platforms can generate antibiotic prescription data and may be

integrated with the pharmacy systems to improve drug ordering [29].

Currently, most public hospitals use electronic systems to monitor stocks on

antiretrovirals, tuberculosis medication and contraceptives [30]. This is because

most of these programmes are supported externally by development agents.

There is, therefore, need to expand these systems to include other standard

treatments and drugs, including antibiotics in the hospitals. Use of electronic

supply chain monitoring can improve the hospital drug procurement process to

ensure stockouts are limited, and drugs do not expire in the stores. These

systems can alert the pharmacist when the stocks are running out and can be

integrated to offer pharmacovigilance support to report adverse drug events as

required by the law [31]. For the clinicians, the platforms with clinical decision

support systems can prompt clinicians on wrong dosages and antibiotic

combinations. Electronic platforms need to be designed with input from clinical

teams who are the end-users. Input from users enhances usability and

adoption, which in turn improves sustainability [32].

344

Future research work

Behavioural research and ethnographic approaches are vital in understanding

the low status of IPC and staff attitudes towards IPC and antibiotic use. I

recommend further work to understand better what drives behaviour and

practice so that such work can help refine future interventions.

Once treatment guidelines are availed, operational, and implementation

research, including human factor engineering studies, are crucial to

understanding what improves guideline use and uptake in our setting. More

research on understanding the patient perspectives on IPC and ABS in resource-

limited settings is also welcome. This is especially key for understating antibiotic

use among the patients in ambulatory settings. Addition of pragmatic

randomised control trials to assess the impact of treatment guidelines on

patient care and outcomes may provide more robust evidence on the role and

value of guidelines on patient care.

Conclusion

Some infection prevention and control and antibiotic stewardship structures

are present in Kenyan hospitals. These are however poorly structured, poorly

led and grossly underfunded. The improvement of IPC and ABS needs to be seen

as part of strengthening the health systems to contain the looming threat of

antimicrobial resistance.

345

References

1. Cox, J.A., et al., Antibiotic stewardship in low- and middle-income countries: the same

but different? Clinical Microbiology and Infection, 2017. 23(11): p. 812-818.

2. Bedoya, G., et al., Observations of infection prevention and control practices in primary

health care, Kenya. Bulletin of the World health Organization, 2017. 95(7): p. 503.

3. Herron, J.B.T., et al., Personal protective equipment and Covid 19- a risk to healthcare

staff? Br J Oral Maxillofac Surg, 2020.

4. Laxminarayan, R., et al., Antibiotic resistance—the need for global solutions. The Lancet

Infectious Diseases, 2013. 13(12): p. 1057-1098.

5. Barasa, E.W., et al., Recentralization within decentralization: County hospital autonomy

under devolution in Kenya. PLOS ONE, 2017. 12(8): p. e0182440.

6. Kohler, J., Fighting corruption in the health sector: methods, tools and good practices.

New York: United Nations Development Programme, 2011.

7. Vian, T., Review of corruption in the health sector: theory, methods and interventions.

Health policy and planning, 2008. 23(2): p. 83-94.

8. Government of Kenya, Public Finance Management Act 2012. 2012, Government

Printer Nairobi.

9. Irimu, G., et al., Tackling health professionals' strikes: an essential part of health system

strengthening in Kenya. BMJ Global Health, 2018. 3(6): p. e001136.

10. Dieleman, M. and J.W. Harnmeijer, Improving health worker performance: in search of

promising practices. Geneva: World Health Organization, 2006: p. 5-34.

11. Gilbert, G.L. and I. Kerridge, The politics and ethics of hospital infection prevention and

control: a qualitative case study of senior clinicians' perceptions of professional and

cultural factors that influence doctors' attitudes and practices in a large Australian

hospital. BMC Health Services Research, 2019. 19(1): p. 212.

12. Shah, N., et al., Towards changing healthcare workers' behaviour: a qualitative study

exploring non-compliance through appraisals of infection prevention and control

practices. Journal of Hospital Infection, 2015. 90(2): p. 126-134.

13. Dixon-Woods, M. and P.J. Pronovost, Patient safety and the problem of many hands.

BMJ Quality &amp; Safety, 2016. 25(7): p. 485-488.

14. yiğit, V., et al., Outsourcing and Its Implications for Hospital Organizations in Turkey.

Journal of health care finance, 2007. 33: p. 86-92.

15. Ministry of Health, The National Antimicrobial Stewardship Guidelines for the Health

Care Settings in Kenya. 2020: Nairobi

346

16. Khalili, H., et al., Role of clinical pharmacists' interventions in detection and prevention

of medication errors in a medical ward. International journal of clinical pharmacy, 2011.

33: p. 281-4.

17. Francis, J. and S. Abraham, Clinical pharmacists: Bridging the gap between patients and

physicians. Saudi Pharmaceutical Journal, 2014. 22(6): p. 600-602.

18. Muhia, J., L. Waithera, and R. Songole, Factors affecting the procurement of

pharmaceutical drugs: A case study of Narok County Referral Hospital, Kenya. Med Clin

Rev, 2017. 3(4): p. 20.

19. Koech, L.C., et al., Quality and Brands of Amoxicillin Formulations in Nairobi, Kenya.

BioMed Research International, 2020. 2020: p. 7091278.

20. Hutchinson, J.M., et al., Measurement of antibiotic consumption: A practical guide to

the use of the Anatomical Thgerapeutic Chemical classification and Definied Daily Dose

system methodology in Canada. The Canadian journal of infectious diseases = Journal

canadien des maladies infectieuses, 2004. 15(1): p. 29-35.

21. GLOBAL-PPS. Global Point Prevalence Survey of Antimicrobial Consumption and

Resistance 2015 [cited 2017 16th February]; Available from: http://www.global-

pps.com/.

22. Sutton, E., L. Brewster, and C. Tarrant, Making infection prevention and control

everyone's business? Hospital staff views on patient involvement. Health Expectations,

2019. 22(4): p. 650-656.

23. Government of Kenya, National Policy on prevention and containment of antimicrobial

resistance 2017: Nairobi.

24. Mutea, N. and D. Cullen, Kenya and distance education: A model to advance graduate

nursing. International Journal of Nursing Practice, 2012. 18(4): p. 417-422.

25. Africa Medical Research Foundation. COVID-19 Training for Health Care Workers

Training. 2020 [cited 2020 18th August ]; Available from:

https://ecampus.amref.org/course/view.php?id=143.

26. Dobkin, C., et al., The Economic Consequences of Hospital Admissions. Am Econ Rev,

2018. 102(2): p. 308-352.

27. Kenya Medical Practitioners and Dentists Board, National training & quality assurance

standards for medical schools & teaching hospitals in kenya. 2015 Nairobi.

28. Smits, H., A. Supachutikul, and K.S. Mate, Hospital accreditation: lessons from low- and

middle-income countries. Globalization and Health, 2014. 10(1): p. 65.

29. Muinga, N., et al., Digital health Systems in Kenyan Public Hospitals: a mixed-methods

survey. BMC Medical Informatics and Decision Making, 2020. 20(1): p. 2.

30. Aywak, D., et al., Pharmacy Practice in Kenya. The Canadian journal of hospital

pharmacy, 2017. 70(6): p. 456-462.

347

31. Ministry of Health Poison and Pharmacy Board. Pharmacovigilance Electronic

Reporting System. 2020 [cited 2020 24th August ]; Available from:

https://pv.pharmacyboardkenya.org/.

32. Muinga, N., et al., Implementing an Open Source Electronic Health Record System in

Kenyan Health Care Facilities: Case Study. JMIR medical informatics, 2018. 6(2): p.

e22-e22.

348

349

Chapter 10

Summary, Samenvatting, Acknowledgements,

Author Portfolio & Resume

350

SUMMARY

Infection prevention and control and antibiotic stewardship: two sides

of the same coin in the prevention of antimicrobial resistance

Infection prevention and control (IPC) and antibiotic stewardship (ABS) are two

major interventions that have been promoted to reduce antimicrobial

resistance. Little is known on the state of IPC and ABS in the Kenyan context,

which is a representation of many low and middle-income countries. Functional

Water sanitation and hygiene (WASH) structures in hospitals are crucial for the

success of IPC in hospitals. This work aimed to assess hospitals Infection

Prevention and Control (IPC) and Antibiotic Stewardship (ABS) capacity as part

of tracking and tackling efforts to limit antimicrobial resistance in Kenya. The

specific aims were;

1. To develop a survey tool that can be applied at the national or sub-

national level to monitor WASH and IPC performance in hospitals.

2. To evaluate the WASH and IPC arrangements in Kenyan county

hospitals and explore how these may vary within a single public health

system.

3. To evaluate the opportunities and challenges for IPC and WASH

implementation in Kenya, in the COVID-19 pandemic.

4. To evaluate antibiotic stewardship arrangements in Kenyan hospitals to

guide action at the regional and national level.

5. To examine patterns of antibiotic use and guideline availability across

hospitals and medical specialities in Kenyan public hospitals.

6. To make a case for further development and use of treatment

guidelines in improving antibiotic usage in hospitals in low and middle-

income countries.

351

7. To examine the feasibility of creating an inpatient data platform to

support the regular assessment of appropriate and correct treatment

in Kenyan hospitals, taking inpatient neonatal units as an example.

The focus on the first part of this work was Infection prevention and control,

encompassing Water sanitation and hygiene.

In chapter 2, this work presents the process of modifying the Water Sanitation

and Hygiene Facility Improvement Tool (WASH- FIT) into a Water Sanitation and

Hygiene facility survey tool (WASH-FAST) at ward and facility level, including

assigning responsibility for action. This process enabled for assessment of 34

WASH indicators at ward level and 65 WASH indicators at facility level in level 4

and 5 public hospitals with performance scores attached to each indicator.

Additionally, three levels of accountability (county government, hospital

management and IPC committees) are identified by the WASH-FAST through a

process that involved different stakeholders. The stakeholders included the

Ministry of Health representatives, hospital infection prevention and control

managers, faculty from the universities and nursing schools and

representatives from non-governmental organisations dealing with IPC and

WASH.

In Chapter 3, a mixed-methods approach which included cross-sectional

surveys and in-depth interviews, was used to assess WASH in 14 public hospitals

with 116 wards. The surveys revealed varying levels of performance within

hospital wards and across the hospitals. The overall hospital performance

ranged between 47 and 71% with IPC and WASH being a low-status activity in

most of the hospitals. In-depth interviews with hospital managers, health

officials from the county governments and frontline health workers were used

to understand the reasons behind these variations. The main findings from

these interviews were the existing differences in the built environment,

352

differences in availability and use of resources within hospitals to improve

WASH and variations in the effort the leadership puts into improving WASH.

Other new insights included the role of outsourcing of cleaning services by the

hospitals to external service providers and how this has helped hospitals focus

on their core mandate of care provision.

In chapter 4, I present how the IPC structures have been affected by the

emergence of COVID-19. Using phone interviews, I carried out 11 interviews

with key hospital leaders in the study hospitals. The pandemic exposed gaps in

the hospitals IPC capacity, including the broken-down hand hygiene and waste

management structures. This caused growing fear and anxiety among the

health care providers who felt exposed to contracting the disease. Conversely,

it led to the IPC committees stepping up and taking up leadership roles to offer

training to clinicians on IPC, including the use of personal protective equipment.

Additionally, in a commentary, I highlight the role played by the infection and

prevention committees in the leadership and accountability for IPC in the

hospitals.

This chapter concludes with a policy brief providing an overview and

recommendations from the WASH/IPC work in Kenya.

The second part of the thesis focussed on antibiotic use and stewardship in

Kenya public hospitals.

Chapter 5 presents the findings from the assessment of the antibiotic

stewardship arrangements. Using a combination surveys and in-depth

interviews, the antibiotic stewardship capacities were assessed from a set of 17

indicators with aggregate scores assigned to each indicator. From these

surveys, only one hospital in the survey had a functional antibiotic stewardship

programme in place. We noted the lack of funds for stewardship activities and

lack of antibiotic use guidelines in the hospitals with aggregate scores of 25 and

353

28% respectively. Five key themes were highlighted from the interviews. These

were leadership for ABS, accountability and expert support, supplies of drugs,

mechanisms for monitoring and reporting, and the policies and practices for

ABS.

Chapter 6 presents findings from data collected on antibiotic use and guideline

availability in the study hospitals. Data on antibiotic use was collected from

3590 hospitalised patients in the 14 hospitals using a point prevalence survey.

From the data, cephalosporins and penicillins were the most frequent

prescriptions. There was limited laboratory use with only two of the

hospitalised patients having bacteriological cultures to inform the antibiotic

therapy. From the survey, in only 53% of the patients was the treatment

deemed appropriate based on the predefined criteria. Treatment guidelines

were only available in the paediatric and neonatal units. The availability of these

treatment guidelines increased the odds of appropriate treatment Odds Ratio

6.44[95% CI 4.81-8.64].

The final part presents some practical solutions to improve antibiotic use and

stewardship in public hospitals.

Chapter 7 of the thesis demonstrates the role played by treatment guidelines

in improving diagnosis and treatment. In this section, I use data and findings

from the antibiotic point prevalence survey to highlight how availability of

guidelines improves treatment appropriateness and why treatment guidelines

need to be extended beyond the paediatric and neonatal units. Here I present

some insights into why policymakers and governments should urgently focus

the guideline development process on the most common diseases. I use the

example of the development of the Kenya basic protocols to present a

successful guideline development process in Kenya.

354

Chapter 8 demonstrates how to improve the use of antibiotics, using an

inexpensive electronic data collection tool and a routine neonatal admission

form. Using routine admission data collected from a neonatal unit in a clinical

information network, this work demonstrates how the review of these routine

data using cycles of audit and feedback improved dosing accuracy of gentamicin

improved overtime for those under 2 kg from 60% to 83% over three years.

355

SAMENVATTING

Infectiepreventie en -bestrijding en antibioticabeheer: twee kanten

van dezelfde medaille bij de preventie van antimicrobiële resistentie

Infectiepreventie en -bestrijding (IPC) en antibiotic stewardship (ABS) zijn twee

belangrijke interventies om antimicrobiële resistentie te verminderen. Er is

weinig bekend over de stand van zaken van IPC en ABS in de Keniaanse context,

die beschouwd kan worden als representatief voor veel lage- en

middeninkomenslanden. Schoon water, sanitaire voorzieningen en hygiene

(WASH) zijn cruciaal voor het succes van IPC in ziekenhuizen. Het onderzoek in

dit proefschrift heeft tot doel om de status van IPC, ABS en WASH in

ziekenhuizen in Kenia te beschrijven en te analyseren, ten behoeve van het

voorkómen en terugdringen van antimicrobiële resistentie. De specifieke

doelstellingen van het onderzoek waren als volgt.

1. Het ontwikkelen van een methode om WASH en IPC in ziekenhuizen te

monitoren die op nationaal of sub-nationaal niveau kan worden toegepast.

2. Het evalueren van de organisatie van WASH- en IPC in Keniaanse provinciale

ziekenhuizen en de variatie hierin binnen één enkel volksgezondheidssysteem.

3. Het beschrijven van de kansen en uitdagingen voor de implementatie van IPC

en WASH in de COVID-19-pandemie.

4. Het evalueren van ABS activiteiten in Keniaanse ziekenhuizen om

maatregelen op regionaal en nationaal niveau te sturen.

5. Het meten van antibioticagebruik en beschikbaarheid van richtlijnen voor

antibioticagebruik in Keniaanse ziekenhuizen.

6. Zich sterk maken voor de verdere ontwikkeling en toepassing van

behandelrichtlijnen voor het verbeteren van antibioticagebruik in ziekenhuizen

in lage- en middeninkomenslanden.

356

7. Onderzoeken van de haalbaarheid van een elektronisch dataplatform voor

de evaluatie van zorg voor intramurale patiënten in Keniaanse ziekenhuizen,

met als casus de afdeling neonatologie.

Het eerste deel van dit proefschrift richt zich op Infectiepreventie en WASH.

Hoofdstuk 2 beschrijft de aanpassing van de Water Sanitation and Hygiene

Facility Improvement Tool (WASH-FIT) naar een Water Sanitation and Hygiene

Facility Improvement Tool (WASH-FAST) dat toegepast wordt op afdeling- en

instellingsniveau, inclusief het toewijzen van verantwoordelijkheden. Deze

aanpassing omvat 34 WASH-indicatoren op afdelingsniveau en 65 WASH-

indicatoren op instellingsniveau ten behoeve van openbare ziekenhuizen

(niveau 4 en 5), met indicatoren waaraan prestatiescores zijn gekoppeld.

WASH-FAST omvat ook drie niveaus van verantwoording, die gezamenlijk zijn

opgesteld met diverse belanghebbenden onder wie vertegenwoordigers van

het Ministerie van Volksgezondheid, managers verantwoordelijk voor

ziekenhuisinfectiepreventie en -bestrijding, vertegenwoordigers van

kennisinstellingen en vertegenwoordigers van niet-gouvernementele

organisaties die zich bezighouden met IPC en WASH.

Hoofdstuk 3 beschrijft een studie, gebruik makend van een enquête en diepte-

interviews, waarin WASH status werd beoordeeld in 14 openbare ziekenhuizen

met 116 afdelingen. Uit de onderzoeken kwam variatie in WASH

prestatieniveaus naar voren tussen ziekenhuisafdelingen en tussen

ziekenhuizen. De algehele scores van de ziekenhuizen varieerden tussen 47%

en 71% van de maximale score, waarbij IPC en WASH in de meeste ziekenhuizen

activiteiten met een lage status bleken te zijn. Diepte-interviews met

ziekenhuismanagers, gezondheidsfunctionarissen van de provinciale

regeringen en eerstelijnsgezondheidswerkers werden gebruikt om deze

variaties te begrijpen. De belangrijkste bevindingen waren de verschillen in de

357

bouwstaat van de ziekenhuizen, verschillen tussen ziekenhuizen in

beschikbaarheid en gebruik van middelen om WASH te verbeteren en

verschillen in aansturing en leiderschap om WASH te verbeteren. Daarnaast

bleek dat de uitbesteding van schoonmaakdiensten aan externe

dienstverleners de ziekenhuizen beter in staat stelde hun kernopdracht van

zorgverlening uit te voeren.

In hoofdstuk 4 presenteer ik hoe de opkomst van COVID-19 van invloed is op

de IPC-structuren. Uit telefonische interviews met belangrijke leiders in de

ziekenhuizen die aan het eerdere onderzoek hebben deelgenomen blijkt dat de

pandemie hiaten in de IPC-capaciteit van de ziekenhuizen bloot legt, waaronder

gebrek aan handhygiëne en afvalbeheer. Deze hiaten veroorzaakten een

groeiende angst op het oplopen van de ziekte bij de zorgverleners die zich

blootgesteld voelden. Omgekeerd leidde het ertoe dat de IPC-comités meer en

meer het initiatief namen om clinici op IPC te trainen, inclusief het gebruik van

persoonlijke beschermingsmiddelen. Daarnaast belicht ik in een commentaar

de rol die de infectiepreventiecommissies spelen bij de leiding en

verantwoording van IPC in de ziekenhuizen.

Dit hoofdstuk wordt afgesloten met een beleidsnota die is gepresenteerd aan

het Ministerie van Volksgezondheid en andere belanghebbenden, met een

overzicht van en aanbevelingen voortkomend uit het WASH / IPC-werk in Kenia.

Het tweede deel van het proefschrift is gericht op antibioticagebruik en ABS in

openbare ziekenhuizen in Kenia.

Hoofdstuk 5 beschrijft de ABS activiteiten in de deelnemende ziekenhuizen aan

de hand van een reeks van 17 indicatoren die zijn getoetst met behulp van

enquêtes en diepte-interviews. Slechts één ziekenhuis rapporteerde een

functioneel ABS programma. Belangrijkste lacunes waren het gebrek aan

financiële middelen voor ABS activiteiten en het ontbreken van richtlijnen voor

358

antibioticagebruik in de ziekenhuizen (totale score van respectievelijk 25% en

28%). In de interviews kwamen vijf hoofdthema's aan bod. Dit waren

leiderschap voor ABS, verantwoordingsplicht en deskundige ondersteuning,

levering van antibiotica, mechanismen voor monitoring en rapportage, en het

beleid en uitvoering van ABS.

Hoofdstuk 6 presenteert bevindingen van het onderzoek naar

antibioticagebruik en de beschikbaarheid van richtlijnen in de deelnemende

ziekenhuizen. Van 3590 gehospitaliseerde patiënten in de 14 ziekenhuizen

werden gegevens over antibioticagebruik verzameld in een puntprevalentie-

onderzoek. Cefalosporines en penicillines bleken het meest voorgeschreven te

worden. Er werd zeer beperkt gebruik gemaakt van laboratorium diagnostiek,

waardoor voor slechts 2 van de gehospitaliseerde patiënten bacteriologische

kweken beschikbaar waren op basis waarvan antibiotica konden worden

voorgeschreven. Uit de enquête bleek dat bij slechts 53% van de patiënten de

behandeling passend werd geacht op basis van de vooraf gedefinieerde criteria.

Behandelrichtlijnen waren alleen beschikbaar op de pediatrische en neonatale

afdelingen. De beschikbaarheid van deze behandelrichtlijnen verhoogde de

kans op een passende behandeling, Odds Ratio 6,44 [95% BI 4,81-8,64].

Het laatste deel van het proefschrift presenteert enkele praktische oplossingen

om antibioticagebruik en ABS in openbare ziekenhuizen te verbeteren.

Hoofdstuk 9 van het proefschrift beschrijft de rol die behandelrichtlijnen spelen

bij het verbeteren van diagnose en behandeling. Dit is vooral belangrijk in

gebieden met beperkte laboratorium- en specialistische ondersteuning. Hier

geef ik een aantal inzichten waarom beleidsmakers en overheden het

richtlijnontwikkelingsproces dringend moeten richten op de meest

voorkomende ziekten. Het voorbeeld van de ontwikkeling van de Keniaanse

359

basisprotocollen in de neonatologie gebruik ik om het

richtlijnontwikkelingsproces in Kenia onder de aandacht te brengen.

Hoofdstuk 8 laat zien hoe het gebruik van antibiotica kan worden verbeterd met

behulp van een goedkope elektronische dataplatform en een standaard

opnameformulier op de afdeling neonatologie. Dit onderzoek laat zien hoe de

beoordeling van routine opnamegegevens met behulp van audit- en

feedbackcycli de nauwkeurigheid van de dosering van gentamicine bij neonaten

onder de 2 kg verbeterde van 60% tot 83% accuraat in een periode van drie

jaar.

360

Acknowledgements

The success of this PhD journey has been made possible by the excellent

support offered to me by several people.

To my promoters and co-promoters, thank you all for reading through

numerous drafts and offering valuable and timely feedback to ensure I give the

best. Prof Mike English; thank you for your patience and always being open to

help and offer direction whenever I needed it. It was a pleasure having you as

a promoter. Prof Constance Schultsz; I have had the honour of working with you

through this PhD, and I enjoyed it. Thank you for always assisting me to frame

the right research questions and for the encouraging me to keep going and keep

my eye on the goal. Dr Olga Tosas Auguet; thank you for the support through

the PhD, your attention to detail came in handy during the project preparation

and execution. The thorough manuscript reviews were critical for me to ensure

I excel. Thank you, Olga! Dr Jacob Mc Knight, your expertise and

encouragement through the PhD is much appreciated. Thank you for always

reminding me to take each day at a time and give my best. You taught me to

keep my head up even when things were not working as I had planned or

hoped. That is a lesson I will carry through life.

To all the co-authors and collaborators on this project, thank you for your time

and professional input to produce this work. I believe the findings from this

work will go a long way in improving infection prevention and antibiotic

stewardship in Kenya and beyond.

Thank you to IDeAL and the KEMRI Wellcome Trust Research Programme for

funding my PhD training and offering professional support through training and

mentorship. For my colleagues and fellow students at the programme, thank

you for always being available to offer your skills whenever I needed assistance.

361

A special thank you to Paul Mwaniki for teaching and supporting me through

the data analysis.

A special mention to my dear wife Nduku Michuki and our wonderful boys

Ethan and Ryan for being patient with me and encouraging me to keep going

even in the tough times. Thank you Nduku for reading through my work and for

ensuring all the commas and full stops were in the right places. Thank you, guys

for the love. This is a win for all of us!

To my parents, brothers and sisters, thank you for the prayers and kind words

through this PhD journey and the immense support through the years.

Without the support of the skilled health workers and hospital staff who work

tirelessly to provide high-quality care in our hospitals, this work would not have

been possible. I want to thank you all for the sacrifices you make to serve our

citizens.

I would also like to appreciate the defense committee for reviewing and

recommending this thesis for defense at the University of Amsterdam.

I may not be able to mention everyone who made this PhD possible; please

receive my sincere gratitude for the support.

To God almighty, thank you for life, opportunity, peace and a sound mind.

362

PhD Portfolio

PhD student: Jackson Michuki Maina

Period: May 2018 – October 2020

PhD supervisors: Prof Constance Schultsz, Prof Mike English, Dr Olga Tosas-

Auguet, Dr Jacob McKnight

1. PhD training

Year Workload

(Hours)

Courses

Writing in the sciences online course hosted by the

University of Stanford. 2018 80

Communication and consenting in research

workshop hosted by the ethics department, KEMRI

Wellcome Trust Research Programme. Kilifi, Kenya.

2018 32

Paediatric Advanced Life Support (PALS) offered by

the department of paediatrics, Aga Khan university

hospital. Nairobi, Kenya.

2018 25

Introduction to qualitative research methods and

data analysis by the KEMRI Wellcome Trust data

team. Nairobi, Kenya.

2018 32

Statistical analysis in R online training hosted by

the data camp group. 2018 100

Implementation science for diseases of online

poverty course offered by the WHO Special

Programme for Research and Training in Tropical

Diseases (TDR).

2018 80

363

Introduction to global health research online

course by the University of Melbourne. 2019 80

Advanced epidemiological analysis at the London

School of Hygiene and Tropical Medicine. London,

UK.

2019 85

Scientific writing and publication course by the

training department at the KEMRI Wellcome Trust

Research Programme. Nairobi, Kenya.

2019 16

Clinical leadership and management offered by the

Oxford executive coaching group. Nairobi, Kenya. 2019 15

Certificate in project management online course by

the University of Washington. 2020 90

Seminars, workshops and masterclasses

Weekly (4 Hrs/Month) scientific seminars and

journal clubs at the KEMRI Wellcome Trust

Research Programme. Nairobi, Kenya.

2018-

2020 144

KEMRI Wellcome Trust Research Programme,

social science group annual workshop. Theme:

Power and intersectionality. Nairobi, Kenya.

2019 15

PhD students day convened at the University of

Oxford. Oxford, UK.

2018

&19 16

KEMRI Wellcome Trust Research Programme, PhD

students, retreat. Kilifi, Kenya. 2019 25

KEMRI Wellcome Trust Research Programme, staff

workshop on disaster management. Nairobi,

Kenya.

2019 8

364

Seminar/ Conference Presentations

Infection Prevention and Control Network annual

scientific conference. Oral presentation on

Infection prevention and control in Kenyan public

hospitals. Nyeri, Kenya.

2018 0.5

Kenya lung conference. oral presentation on

respiratory infections in children. Nairobi, Kenya. 2018 0.5

Webinar presentation on Water Sanitation and

Hygiene in Kenya hosted by the WHO WASH

programme in Geneva, Switzerland.

2018 1

PhD Pre-registration seminar presentation hosted

at KEMRI Wellcome Trust Research Programme.

Nairobi, Kenya.

2019 1

PhD students day presentation on WASH and

antibiotic use in Kenyan hospitals hosted at the

University of Oxford. Oxford, UK.

2018

&19 2

Seminar presentation at the KEMRI Wellcome

Trust Research Programme, Title: Water Sanitation

and Hygiene in Kenya. Nairobi, Kenya.

2019 1

Oral presentation on Infection Prevention and

Control in Kenya at the Ministry of Health, quality

of care report launch. Nairobi, Kenya

2019 1

Oral presentation to the Ministry of Health

Infection Prevention and Control technical working

group. Nairobi, Kenya.

2019 3

365

Oral presentation/ webinar at the WHO quality of

care network, Geneva, Switzerland.

Webinar link:

http://www.qualityofcarenetwork.org/webinars/re

cording-and-materials-webinar-wash-quality-care-

kenyan-hospitals.

2020 1

(Inter)national conferences

Oral presentation and attendance at the crossing

boundaries conference at the University of Oxford.

Oxford, UK.

2018 15

Attended the Kenya Paediatric Association (KPA)

annual scientific conference and facilitated

breakout sessions on research in paediatrics.

Mombasa, Kenya.

2018 35

Poster presentation African Academy of Science

(AAS) international conference. Dakar, Senegal. 2019 25

Risk factors for neonatal sepsis in low resource

settings stakeholders technical meeting hosted by

the Centers for Disease Control (CDC). Atlanta,

USA.

2019 25

KEMRI annual scientific conference was in

attendance and gave an oral presentation on

Infection Prevention and Control strategies in

Kenyan hospitals. Nairobi, Kenya.

2020 25

Others – Policy Engagement

366

Member and attendee of the Ministry of Health,

Infection Prevention and Control technical working

group meetings.

2018

&19 18

Attended a meeting hosted by the Ministry of

Health to discuss strategies for sensitization of

health workers on antimicrobial resistance.

Nairobi, Kenya.

2019 8

2. Teaching Year Hours

Lecturing

Diploma in Tropical Medicine and Hygiene

(DTM&H). London School of Hygiene and Tropical

Medicine.

(Teaching ETAT+ and Newborn Care Scenarios).

Kampala, Uganda.

2017 &

19

40

The University of Nairobi, The Partnership for

Health Research Training in Kenya (P-HERT);

Systematic review course conducted yearly in

March-July. Nairobi, Kenya.

2018

&19 12

Faculty providing training for health care workers

on a programme to improve the safe use of oxygen

in public hospitals in Kenya under the Clinton

Health Access Initiative (CHAI). Nairobi, Kenya.

2019 &

20 122

Teaching University of Nairobi, paediatric residents

and faculty on systematic reviews on the weekly

journal club. Online teaching series in April 2020.

2020 2

367

Emergency Triage and Treatment (ETAT+) course

faculty. Nairobi, Kenya. 2019 25

Kenya Medical and Dentists Practitioners Council

(KMPDC), external examiner

(6 Days /Year). Nairobi, Kenya.

2018-

2020 50

3. Parameters of Esteem

Year

Awards and Prizes

Appointed into the Kenya Ministry of Health technical working

group on Infection Prevention and Control.

Best oral presentation KEMRI annual scientific conference.

Nairobi Kenya.

2019

2020

Leadership

PhD students representative- KEMRI Wellcome Trust

Programme. Nairobi, Kenya.

2019

4. Publications

Kebaya LMN, Kiruja J, Maina M, Kimani S, Kerubo C, McArthur

A, et al. Basic newborn resuscitation guidelines for healthcare

providers in Maragua District Hospital: a best practice

implementation project. JBI database of systematic reviews

and implementation reports. 2018;16(7):1564-81.

Maina M, Aluvaala J, Mwaniki P, Tosas-Auguet O, Mutinda C,

Maina B, et al. Using a common data platform to facilitate

audit and feedback on the quality of hospital care provided to

sick newborns in Kenya. BMJ Glob Health. 2018;3(5): e001027.

2018

368

Morgan MC, Maina B, Waiyego M, Mutinda C, Aluvaala J,

Maina M, et al. Pulse oximetry values of neonates admitted

for care and receiving routine oxygen therapy at a resource-

limited hospital in Kenya. Journal of paediatrics and child

health. 2018;54(3):260-6.

Maina M, Tosas-Auguet O, McKnight J, Zosi M, Kimemia G,

Mwaniki P, et al. Extending the use of the World Health

Organisations' water sanitation and hygiene assessment tool

for surveys in hospitals - from WASH-FIT to WASH-FAST. PLoS

One. 2019;14(12): e0226548.

Maina M, Tosas-Auguet O, McKnight J, Zosi M, Kimemia G,

Mwaniki P, et al. Evaluating the foundations that help avert

antimicrobial resistance: Performance of essential water

sanitation and hygiene functions in hospitals and

requirements for action in Kenya. PLoS One. 2019;14(10):

e0222922.

McKnight J, Maina M, Zosi M, Kimemia G, Onyango T, Schultsz

C, et al. Evaluating hospital performance in antibiotic

stewardship to guide action at national and local levels in a

lower-middle-income setting. Global health action.

2019;12(sup1):1761657.

Roberts DJ, Njuguna HN, Fields B, Fligner CL, Maina J, Zaki SR,

Keating MK, et al. Comparison of Minimally Invasive Tissue

Sampling with Conventional Autopsy to Detect Pulmonary

Pathology Among Respiratory Deaths in a Resource-Limited

2019

369

Setting. American journal of clinical pathology.

2019;152(1):36-49.

Njuguna HN, Zaki SR, Roberts DJ, Fligner CL, Maina J Keating

MK, Rogena E, et al. Determining the Cause of Death Among

Children Hospitalized with Respiratory Illness in Kenya:

Protocol for Pediatric Respiratory Etiology Surveillance Study

(PRESS). JMIR Res Protoc [Internet]. 2019 2019/01//; 8(1):

[e10854 p.]

Maina M, Mwaniki P, Odira E, Kiko N, McKnight J, Schultsz C,

et al. Antibiotic use in Kenyan public hospitals: Prevalence,

appropriateness and link to guideline availability.

International Journal of infectious diseases 2020. 99: p. 10-18.

Maina M, Tosas-Auguet O, English M, Schultsz C, McKnight J.

COVID-19: an opportunity to improve infection prevention

and control in LMICs. The Lancet Global Health

Maina M, Tosas-Auguet O, English M et al. Infection

prevention and control during the COVID-19 pandemic:

challenges and opportunities for Kenyan public hospitals

[version 1; peer review: 1 approved]. Wellcome Open

Res 2020, 5:211

(https://doi.org/10.12688/wellcomeopenres.16222.1)

2020

Other Publications

370

Involved in the delphi process for the World Health

Organization "Standards for improving the quality of care for

children and young adolescents in health facilities."

https://apps.who.int/iris/bitstream/handle/10665/272346/97

89241565554-eng.pdf

Part of PhD work included in a report by the World Health

Organization: Water, sanitation and hygiene in health care

facilities: practical steps to achieve universal access.

https://apps.who.int/iris/bitstream/handle/10665/311618/97

89241515511-eng.pdf

Member of the WHO technical working group that is updating

the water and sanitation assessment tools for health facilities.

2020

2018

2019

2020

371

RESUME

Biodata

Name: Dr Jackson Michuki Maina

Date of Birth: 29th July 1982

Personal Statement

Michuki is committed to professional and academic excellence that enables

him to participate in and undertake health research and teaching to improve

newborn, child and adolescent health in Kenya.

Academic Qualifications

Qualification Institution Year

PhD Candidate University of Amsterdam 2018 – Present

Master of Medicine (MMed) in Paediatrics and child health

University of Nairobi 2011-2014

Master of Public Health (MPH)

Moi University/ African Medical Research Foundation (AMREF) 2010-2014

Bachelor of Medicine and Surgery (M.B Ch.B.)

University of Nairobi, September

2002-2007

Professional Experience

Role Institution Year Research paediatrician/ PhD fellow

KEMRI/Wellcome Trust Research Programme

2014- Present

Consultant paediatrician

Maria Immaculate Hospital/ AAR Healthcare

2014-Present

Honorary consultant paediatrician

Kenyatta National Hospital 2015- 2017

Medical doctor AAR Healthcare November 2009-July 2014

Medical officer Ministry of Health- Embu Level 5 Hospital

January 2009 –October 2009

372

Additional Training

Instructor Emergency Treatment and Triage plus Admission (ETAT+), March

2015.

Paediatric Advanced Life Support (PALS), February 2018.

Qualitative research methods training KEMRI Wellcome Trust Research

Programme. February 2018.

Scientific writing and publication Course KEMRI Wellcome Trust Research

Programme February 2018.

Statistical analysis in R online training May 2018.

Introduction to global health research. University of Melbourne (Online 8

weeks course).

Advanced epidemiological analysis, London School of Hygiene and Tropical

Medicine 9th -20th September 2019.

Implementation science training by the World Health Organization 2019 (8

weeks).

University of Washington project management for global health online 11-

week course January – March 2020.

Teaching Experience

Part of the faculty that provides teaching on systematic reviews to faculty and

postgraduate students at the University of Nairobi school of medicine under

The Partnership for Health Research Training in Kenya (P-HERT) (Course

conducted yearly in March-July from 2017).

Faculty providing training for health care workers on a programme to improve

the safe use of oxygen in public hospitals in Kenya under the Clinton Health

Access Initiative ( October- November 2019).

Teaching on the London School of Hygiene and tropical medicine East Africa

Diploma in Tropical medicine in Uganda course in November since 2017.

Course director on the generic instructors course for the Emergency Triage and

Treatment Programme December 2017 – 2019.

Teaching paediatric residents and faculty on systematic reviews on the weekly journal

club. April 2020.

373

Research grants awarded

Initiative to Develop Research Leaders in Africa. Grant award for PhD training

amount GBP 100,000 by the Africa Academy of Science. October 2017.

Roles and Awards

Hillman award for outstanding leadership skills (the University of Nairobi,

Department of Paediatrics and Child Health 2014.

WHO global consultant on paediatric quality of care 2017.

Member on the national Ministry of Health Infection Prevention and Control

technical working group 2018- Present.

Kenya Medical and Dentists Practitioners Council external examiner -2015-

Present.

Best oral presentation KEMRI annual scientific conference February 2020.

Membership and Affiliations

Registered medical practitioner and consultant paediatrician by the medical

practitioners and dentist’s council Kenya.

Member Kenya paediatric association.

Member of the European Society of Paediatric Infectious Diseases (ESPID).

Medical Missions Africa- Volunteer paediatrician.

Interests and hobbies

Music- Guitar playing and sport-recreational runner.

374