Systematic review on implementation of mobile health (mHealth) projects in Africa: What works? What doesn’t work and why?

Clara Aranda-Jan1, Dr. Neo M. Mohutsiwa-Dibe2, Dr. med. Svetla Loukanova3 (1) Institute for Manufacturing, University of Cambridge, Cambridge, UK. (2) P.O. Box 80912, Gaborone, Botswana. (3) Institute of Public Health, University of Heidelberg, Heidelberg, Germany

     

INTRODUCTION Access to mobile phone technology has rapidly expanded in developing countries. In Africa, mHealth is a relatively new concept and questions arise regarding reliability of the technology used for health outcomes. The main objective of this study is to analyze the experience of implementing mHealth projects in Africa during the last decade, and identifies factors influencing the successes and failures of mHealth projects in Africa using a SWOT (strengths, weaknesses, opportunities and threats) analysis.

RESEARCH QUESTIONS a) What are the factors leading to successful implementation of mHealth projects? b) What are the factors limiting or challenging the implementation of mHealth projects? c) Why do these factors cause project failure or limit project implementation?

RESULTS AND DISCUSSION

Fig. 1 Inclusion/exclusion flowchart

The results from this study on mHealth projects in Africa sought to answer, “what is working with regards to improving population health? what is not working, and why?” mHealth implementations have potential to become an important part of the health sector to establishing innovative approaches to delivering care and benefits have been highly praised, but is clear that mHealth projects are not a solution to the challenges that health systems face in many African countries. Evidence remains poor, results are still project- or setting-specific and questions regarding impact, scalability, increase coverage (e.g. different diseases, different settings, different target populations), cost-effectiveness and sustainability of the projects in Africa are yet to be addressed. While mobile phone technology continues to improve, more research on these areas is essential to fully understand the potential of these projects and help to reach the hard isolated and marginalized communities in low and middle income countries (LMICs).

CONCLUSIONS

Table 1. SWOT analysis of included studies The table below summarizes and compiles findings of the studies included in the review. Results in bold to highlight their frequency of appearance in the studies (Table adapted).

METHODS An electronic systematic literature search was conducted using PubMed and Journal @Ovid. Two search strategies were performed: firstly, we used the combination of the MESH terms “mHealth” AND “Africa”, and secondly, we combined the free-text words “mobile phone$ or cellphone$” AND “health” AND “Africa”. The searches were limited for the period between 2003-2013 and included full text articles in English. Potential abstracts were screened and studies were selected for full-text review. Exclusion criteria were: Project not located in Africa, non-mHealth implementation (telemedicine, other types of eHealth and use of other telecommunication technologies, such computer, internet or e-mail), and studies on factors associated to mobile phones but not mHealth implementation (e.g. community ownership or acceptability of mobile phones). Except for project protocols, all study designs (randomized-control trials [RCTs], pilot project, literature reviews etc.) were included.

!"#$%&' ()"*+,-./012&'

!"#$%&'&()*+

343'!"#$%&'&()*+

533'

366'#,*)$#%)*'"-%./0"0',#*"0'

12'(2%./*(123

"-%./*(12'

%$()"$(#'!"."%)"0'41$'

4/..5)"-)'

$"6("77'65'

88'2+9-"&%&'.:"&2%.'

;<'"-%./0"0',#*"0'12'

(2%./*(123

"-%./*(12'

%$()"$(#'

8$/9'*/::.;'#20'

*)1%<'

=#2#9"="2)+';'

>#?"2)'"0/%#?12'

#20'#7#$"2"**+'5'

>(.1)'*)/0;+';' >(.1)'*)/0;+'5'

>#?"2)'41..175/:'#20'

="0(%#?12'

#0&"$"2%"+'5='

>(.1)':$1@"%)+'>'ABC*+'4'D()"$#)/$"'$"6("7+';'E(-"0'=")&10*+';'B$1**5*"%?12#.'

*)/0;+'5'B1&1$)'*)/0;+'5'

!)#F'"6#./#?12G'

=12()1$(29'#20'

9/(0".(2"*H'

%1=:.(#2%"+'8'

ABC+';'E(-"0'=")&10*+'5'B1*)5#2#.;*(*+'I'

J6"$6("7'

=K"#.)&'

:$1@"%)*+'';'

D()"$#)/$"'$"6("7+'5'L/#.()#?6"'*)/0;+'5'

<?>''C1)#.'&()*'

M,*)$#%)*')1'

,"'$"6("77'8>8'

8/:.(%#)"*'".(=(2#)"0+'

34'/*(29'N2021)"*G'<'=#2/#..;O'

5;'4/..5)"-)'21)'

#6#(.#,."'

<'#,*)$#%)*'#00"0'

,#*"0'12'

$"4"$"2%"5

.(*)*'$"6("7'

<'*)/0;':$1)1%1.*'3'21)'M4$(%#'P"-%./*(6".;Q'55'21)'=K"#.)&'(=:."="2)#?12*'

>''9"2"$#.'#*:"%)*'14'=1,(."':&12"*'

!)#F')$#(2(29G'

*/::1$)'#20'

=1?6#?12+'R'

>(.1)'*)/0;+';'

8(*"#*"'

*/$6"(..#2%"'#20'

(2)"$6"2?12'

=12()1$(29+'8'

>(.1)'*)/0;+';'B$1**5*"%?12#.'

*)/0;+'5'B1&1$)'*)/0;+'5'

8#)#'%1.."%?123

)$#2*4"$'#20'

$":1$?29+'5?'

>(.1)'*)/0;+'>'ABC+'5'E(-"0'=")&10*+'5'B$1**5*"%?12#.'

*)/0;+'5'B1*)5#2#.;*(*+'5'

• High  availability  of  mobile  phones  • Overall  posi6ve  percep6on  and  poten6al  of  posi6ve  outcomes  at  small-­‐scale  • Adapta6on  to  local  context  • Provision  of  incen6ves  • Not  resource  exhaus6ve  and  low  replica6on  costs  

Successful  factors  

• Unclear  long-­‐term  results  and  benefits  • Scarce  large-­‐scale  studies  • Risk  of  biased  repor6ng  and  weak  evidence  • Unknown  cost-­‐effec6veness  • Lack  of  integra6on  to  the  health  system  • Poor  evidence  on  standards,  privacy  and  legal  issues  

Unsuccessful  factors  

• Lack  of  adequate  planning  and  poor  project  design  • Limited  funding  for  long-­‐term  projects  • Research  limited  to  pilot-­‐project  or  donor  repor6ng  • External  factors:  culture,  illiteracy,  treatment  dura6on.  • Unclear  roles  and  responsibili6es  in  government  and  ministries;  lack  of  an  organiza6onal  capacity.  Lack  of  standards,  guideline,  policies  and  regula6ons  

• Limited  local  technical  support  and  capacity  

Reasons  for  failure  

Fig. 2 What works? What does not work? And, why?

Factors Strengths Weaknesses Opportunities Threats

Mid- and long-term results/

Project sustainability

- Improve delivery of services (e.g. skilled delivery attendance) and service request (e.g. appointments) - Increased health workers’ adherence to clinical guidelines and quality of treatment, worker morale and sense of empowerment, access to medical/health information at the point-of-care, and motivation due to training and improved skills - Overcome communication delays, ensure real-time data acquisition and reporting, reduces data losses and monitor data quality, makes available pre-define indicators and reduces delayed reporting - Decreases referral time and care costs burden to patients due to transportation - Supports disease surveillance systems and monitoring of interventions

- Unclear benefits, uncertain long-term results and effectiveness (e.g. insufficient results from RCTs), and unclear cost-benefit analysis. - Results are variable depending on the duration of the intervention and may be overestimated, limited study design and external validity, weak evidences - Difficult to monitor text messages content, high possibility of data under-reporting, and possibility of biased responses from participants - Reported patient anxiety due receiving information

- Potential to enhance timeliness in reporting health and stock data in rural and remote areas - Lack of stock management resulting in patients untreated - mHealth projects are regarded as innovative and current data collection methods tend to have poor quality

- High facility workload and staff/patient/user illiteracy - Limited knowledge on the effects of mHealth on patient health outcomes in low-resource settings - Use of mobile technology for research is recent - Dependency in donor funding and limited funding opportunities may limit long-term sustainability - mHealth results are dependant of external factors (e.g. long duration of patient treatment may reduce adherence and motivation to participate) - Costs of mHealth implementation may affect patient treatment costs

Integration into the health

system

- Support patient management - Intervention flexible to be adapted to local context and language - Allows focusing efforts of clinical staff in areas not covered by the intervention - Public-private partnerships proved to work effectively in these projects - High government commitment, existing governmental eHealth strategy - Availability of local private providers willing to set up the mHealth system

- Unclear roles, responsibilities, actions, boundaries and responses needed at different levels of healthcare system (government) for project implementation and scale-up - Project results depend on training and clinical practice of health workers - Most pilot projects are started by implementing organisations themselves rather than integrated to the health system - mHealth projects are unlikely to prove effective in poorly performing systems

- Poor management of drug supply chain and large discrepancies of and limited control in stock levels of health facilities, and poor stock forecasting - Opportunities to be implemented in different national disease control programmes; provide access to data for an evidence-based approach - Project may be attractive and acceptable for private or commercial partners and governments (MoH)

- Political crisis may hindered project implementation and results - Current care delivery processes will need to be redesigned (e.g. change to electronic records and data) -Unknown health systems complexities for large scale implementation of mHealth projects - Lack of cultural and organisational capacity to manage digital health information may lead to late reporting, lack of feedback and incomplete data collection

Project management

process

- Support provision of user and staff training - Minimal human resources and training are required - Financial incentive (e.g. airtime credit) allows high response rate to the project - Allows real-time supervision and monitoring work rate, attendance, and staff working hours

- Low patient motivation to participate when project is not tailored to their needs (e.g. local language) - Small sample size of pilot projects provide limited or biased results - Occasional staff shortages during project implementation, and staff may be overwhelmed of increased calls or messages

- Available funding from larger programmes (e.g. PEPFAR mobile clinic) - Low capacity and administrative challenges for data collection

- Challenge of management of mHealth projects remain underestimated

Legal issues, regulations

and standards

- Coded information contributes to data security and confidentiality - Integration of SMS guidelines into healthcare process delivery

- Privacy concerns raised when using mobile phones, particularly if not owned by the patient

Not mentioned

- No minimum number of critical surveillance parameters to be reported has been established - Lack of published data on feasibility and acceptability of confidentiality methods

Technology and

infrastructure

!"Text messaging is inexpensive, uses existing infrastructure (e.g. existing networks, reducing phone costs) "

!Users are familiar to mobile phone services "

- High acceptance, satisfaction and valued by patients and staff

- Technical challenges reduce data quality and transfer, lost network, phone maintenance costs and risk of theft

- High access and rapid expansion of mobile network coverage, availability of inexpensive handsets, and decreasing costs of mobile phone services and rapidly-growing technological field

- Dependency on network coverage - High illiteracy and users’ preference makes voice calls more attractive than text messaging - Unreliable network, internet and electricity access

Scale-up and replication

- Low replication costs and highly adaptable to specific cultural contexts - High potential to be scaled-up

- No assessment has been performed to know if an effective implementation for one disease works for other diseases .

- Cost-effective implementation of m-Health programmes (e.g. lower running costs)

- Unknown cost-effectiveness of deployment and maintenance

!