Upload
others
View
1
Download
0
Embed Size (px)
Citation preview
Guidelines on systematic screening for active TB
Knut Lönnroth, WHO/STB/PSI The Wolfheze meeting,
The Hague, 30 May 2013
1
• The recommendations in this guideline were developed by a Guideline Development Group consisting of the following experts: Dr Sevim Ahmedov (United States Agency for International Development, United States), Dr Helen Ayles (University of Zambia, Zambia), Dr Lucie Blok (KIT Royal Tropical Institute, the Netherlands), Dr Gavin Churchyard (Aurum Institute for Health Research, South Africa), Dr Liz Corbett (MLW Clinical Research Programme, Malawi), Dr Mao Tang Eang (National TB Programme, Cambodia), Dr Peter Godfrey-Faussett (London School of Hygiene and Tropical Medicine, United Kingdom), Dr Jonathan Golub (Johns Hopkins University School of Medicine, United States), Dr Katharina Kranzer (London School of Hygiene and Tropical Medicine, United Kingdom), Dr Josué Lima (National TB Programme, Brazil), Dr Wang Lixia (National TB Programme, People’s Republic of China), Dr Thandar Lwin (National TB Programme, Myanmar), Dr Ellen Mitchell (KNCV Tuberculosis Foundation, the Netherlands), Dr Mary Reichler (Centers for Disease Control and Prevention, United States), Dr Adrienne Shapiro (John Hopkins University School of Medicine, United States), Dr Alena Skrahina (National TB Programme, Belarus), Dr Pedro Guillermo Suarez (Management Science for Health, United States), Dr Marieke van der Werf (European Centre for Disease Prevention and Control, Sweden), Dr Anja Van’t Hoog (Amsterdam Institute for Global Health and Development, the Netherlands), Dr Norio Yamada (Research Institute for TB, Japan). Dr David Sinclair (Cochrane Infectious Diseases Group and Liverpool School of Tropical Medicine, United Kingdom) served as the guideline’s methodologist.
• The WHO steering group consisted of Dr Haileyesus Getahun, Dr Knut Lönnroth, Dr Ikushi Onozaki, Dr Salah Ottmani, Dr Mario Raviglione, Dr Mukund Uplekar, and Ms Diana Weil, from the Stop TB Department, and Mr Jacob Creswell and Dr Suvanand Sahu from the Stop TB Partnership secretariat. In addition, the following WHO staff contributed to the guideline’s development: Dr Philippe Glaziou, Dr Malgosia Grzemska, Dr Dennis Falzon, Mr Wayne Van Gemert and Dr Karin Weyer from the Stop TB Department; Dr Andreas Reis from the Department of Ethics, Equity, Trade and Human Rights; Dr Daniel Kibuga from the Regional Office for Africa; and Dr Catharina van Weezenbeek from the Regional Office for the Western Pacific. The development of the guideline was coordinated by Dr Knut Lönnroth, who also wrote the first draft. All members of the Guideline Development Group and WHO Secretariat staff involved in developing the guideline reviewed and commented on several iterations of the draft.
• Peer review comments on the prefinal draft were received from Dr Martien Borgdorff (University of Amsterdam, the Netherlands), Dr Frank Cobelens (Amsterdam Institute for Global Health and Development, the Netherlands), Dr Paul Douglas (Department of Immigration and Citizenship, Australia), Dr Steven Graham (Childhood TB Subgroup, Centre for International Child Health, Australia), Dr Anthony D Harries (International Union Against Tuberculosis and Lung Disease, United Kingdom), Dr Giovanni B Migliori (European Respiratory Society, WHO Collaborating Centre for TB and Lung Diseases, Italy), Dr Lisa Nelson (WHO HIV Department), and Dr Alasdair Reid (UNAIDS, South Africa).
• Funding through TB CARE (USAID)
Acknowledgements
2
1. Rationale, definitions and objectives
2. Summary of systematic review findings
3. Recommendations on risk groups to screen
4. Screening and diagnostic algorithm options
5. Key principles
6. Need for tools and operational guidelines
Presentation overview
3
Definition
"Systematic screening for active TB" Systematic identification of people with suspected active TB, in a predetermined target group, by the use of tests, examinations, or other procedures which can
be applied rapidly
Among those screened positive, the diagnosis should be established through diagnostic tests and clinical assessments with combined high sensitivity and specificity.
Can target: a) people who do not seek care (access barriers, not recognising symptoms as
serious, etc) or
a) people who seek care (with or without symptoms/signs compatible with TB), e.g.
specific clinical risk groups
4
• The primary objective of screening is to detect active TB early; this contributes to two ultimate goals: 1. reducing the risk of poor treatment outcomes, health sequelae, and
the adverse social and economic consequences of TB for the individual. This can help reduce suffering, the prevalence of TB, and death from TB;
2. reducing TB transmission by shortening of the duration of infectiousness. This can help reduce the incidence of TB.
• A second objective is to rule out active disease to help identify people
who are eligible for treatment of latent TB infection – for example, among people living with HIV and contacts who are younger than 5 years.
Screening objectives & goals
5
Σ (individual health + externalities) = public health
Active TB
Patient recognizing symptoms
Patient accessing
health care
Notification & treatment
Improving knowledge
& awareness
Improve referral and notification
practices
Ensuring quality-assured diagnosis
Infected
Patient delay
Suspect identification
delay
Access delay
Strengthening identification of patients
with suspected TB
Contacts -Children
-Other risk groups
-All household
-Workplace
-Wider
Clinical risk groups -HIV
-Previous TB
-Malnourished
-Smokers
-Diabetics
-Drug abusers
Risk populations -Prisons
-Urban slums
-Poor areas
-Migrants
-Workplace
-Elderly/infants
The screening pathway
Minimize barriers to health care access
Identification of patient requiring
TB test
Completing high quality
diagnosis
Exposed
Natural history delay
Diagnostic delay
Treatment delay
The patient-initiated pathway
Improving early detection
Why a screening guideline?
• What we knew: Many people in the community have undetected TB Many of them have vague symptoms Long health seeking and health system delays common
Burden of undetected TB often concentrated to some risk groups Screening can identify TB in symptomatic and non-symptomatic persons Screening can be very expensive and huge numbers of people need to be
screened if prevalence is moderate to low and indiscriminate screening is done
• What the guideline should answer:
? Can screening in risk groups help improve early TB detection, and reduce morbidity and mortality?
? Is it cost effective compared to other interventions? ? When is screening appropriate? ? What risk groups should be prioritized? ? What screening/diagnostic tools/algorithms should be used? ? What key principles should apply?
7
Passive case finding seems insufficient
WHO expert committee on TB, 1974:
“…the policy of indiscriminate tuberculosis case-finding by mobile mass radiography should now be abandoned.”
Guideline development process • Scoping meeting 2011
• Four systematic reviews commissioned:
the general benefits of TB screening (Review 1); the sensitivity and specificity of different screening tools
(Review 2); the number needed to screen to detect one case of TB in
different risk groups (Review 3); the acceptability of screening in different risk groups (Review 4).
• Final guideline meeting October 2012
• Peer-review November- December 2012 (including
symposium in Kuala Lumpur and presentations in various meetings)
• WHO Guideline Review Committee approval March 2013
8
Summary of results from the systematic reviews
9
Review 1: Comparator and key outcomes
10
Screening (overall and by risk group)
in addition to passive case
finding
Vs.
Passive case finding alone
Important outcomes
Change in case detection
Time from onset to diagnosis
Severity of disease at diagnosis
Critical outcomes:
Treatment outcomes
Social/economic consequences
TB burden in the community
Intervention Impact
Impact on transmission/population burden
Important (intermediary) outcomes •Low quality evidence that screening can improve case detection (if large number of people screened) •Low quality evidence that screening can reduce delay to diagnosis
Critical outcomes •Moderate quality evidence that community screening with modest intensity and sensitivity has no impact on TB transmission or TB burden in the community •Low quality evidence that more intense and sensitive screening impacts on transmission within and outside the risk group (=impact uncertain)
Conclusions: 1.Don't expect impact on transmission from community screening of low/moderate intensity
2.More research needed to assess epidemiological impact of more intense screening, in the community and within targeted risk groups
3.Until better evidence is available, the primary goal of screening should be to improve health outcomes for the individual
11
Individual level impact
Direct evidence of very low/low quality (review 1): Important outcomes •Screening can improve case detection, reduce delay and identify people with less severe disease
Critical outcomes •Treatment outcomes are the same for screened and "passively" detected. •No studies on impact on quality of life, social or economic consequences (positive or negative) •No empirical studies assessing relative cost-effectiveness of screening
12
Individual level impact, cont
Indirect evidence of potential positive impact (review 3-4 and other data) •Burden of undetected TB is high in certain risk groups – and many can be detected through screening •TB treatment reduces suffering and death dramatically (vs. no treatment) •Some risk groups also have high risk of poor TB treatment outcomes, which probably can be lowered with early detection and treatment •Acceptability/consent rate is generally high •Some indirect cost data;
Many interventions costly, labor-intense, and very high NNS Some screening interventions easy to do and with low NNS
13
Individual level impact, cont
Indirect evidence of adverse impact (review 2 and other data): •Many false positive cases when TB prevalence is low and the screening and diagnostic algorithm has diagnostic specificity <100%
Problem already when prevalance is less than 1% (1,000/100,000)
•Side effects of treatment •High direct and indirect cost to patients •Stigma and discrimination
14
Individual level impact, cont.
Conclusions: 1. Screening can reduce suffering from TB, especially in vulnerable groups
2. Screening can be very resource demanding
3. Screening can cause harm to individuals
Therefore: Prioritize groups with high likelihood of individual benefit and low risk of harm:
a) High TB prevalence b) High risk of severe negative consequences if diagnosis is missed/delayed c) Poor access to general health services d) Low risk of harm from screening e) Screening is feasible and affordable
15
NNS in medium and high incidence settings
16
13 25
110
35 20
111
269
506
603
120
10 17 36
7
795
100
0
100
200
300
400
500
600
700
800
Medium Incidence
High Incidence
1,200
Health facility "Community"
Groups with Weighted mean ≥80%
Risk Groups (#) Mean
% weighted. mean
% SD %
Min %
Max %
Median %
Military 96 99 0.05 93 100 96
Farm & Factory (5) 97 97 0.02 95 98 98
Homeless (5) 66 96 0.24 41 97 75
Mentally Ill (2) 94 95 0.02 93 95 94
Diabetics (2) 96 94 0.03 94 98 96
Pregnancy (3) 85 94 0.12 68 96 90
Drug Dependent (5) 84 93 0.13 69 94 89
Urban poor areas (11) 87 91 0.12 59 99 88
Sex Worker(2) 86 84 0.12 84 88 88
Children Under 5 (2) 87 87 84 91
Transport workers (3) 85 86 0.13 73 98 84
Migrants (8) 77 85 0.15 55 96 81
PMTCT (3) 79 81
0.14 68 96 88
Immigration 73 80 0.20 34 96 79
Source: Mitchell 2012
Groups with Weighted mean <80%
Risk Groups (#) Mean
%
weighted. mean %
SD %
Min %
Max %
Median %
Transgenders* 77 79 0.07 77 91 84
PLHIV 82
78
0.13 52 99 83
Elderly- Institutionalized * 83
72
0.17 72 95 83
Prisoners 71
72
0.29 18 98 86
Miners 81
70
0.11 66 93 84
Source: Mitchell 2012
Groups with Weighted mean <80%
Risk Groups (#) Mean
% weighted. mean %
SD %
Min %
Max %
Median %
Indigenous 82
69
0.20 40 97 89
Persons testing for HIV (VCT) 73
69
0.23 41 97 85
Attendees of Health centers* 59
68
0.07 52 67 57
Men who have sex with men (MSM)* 76
61
0.48 64 91 -
Health Care Workers (HCW)* 78
59
0.14 56 91 80
Source: Mitchell 2012
Contacts of an infectious TB Patient
Type of Contact Tracing
Mean % N S.D.
Min %
Max %
Median %
Contract Tracing- Household
80 24 0.17 39 99 85
Contact Tracing- Community
88 27 0.12 57 100 91
Contact Tracing- Health care
59 4 0.14 43 73 60
Source: Mitchell 2012
Acceptability of TB screening: summary
• High participation in almost all groups when screening is designed for them.
• Screening: The faster and least invasive the better (1 step process) for hard to reach/hard to keep group
• Who does the screening matters
• Where it takes place matters
• HIV testing may deter uptake of screening in some risk groups, however inclusion of HIV testing in TB screening was not found to increase refusal in community-based TB screening
• Acceptability of screening vs. diagnosis vs. treatment
Recommendations on prioritization of
risk groups to screen
22
Priority?
Feasibly and cost
Benefits Harm
Expert opinions
Evidence
First, key principle No 2:
• Indiscriminate mass screening should be avoided.
• The prioritization of risk groups for screening should be based on assessments made for each risk group of; the potential benefits and harms; the feasibility of the initiative; the acceptability of the approach; the number needed to screen; and the cost effectiveness of screening.
23
Strong recommendations = Should be screened in all settings
1. Household contacts and other close contacts should be systematically screened for active TB.
2. People living with HIV should be systematically screened for active TB at each visit to a health facility.
3. Systematic screening for active TB should be done in current and former workers in workplaces with silica exposure
24
Conditional recommendations = prioritization needed
4. Systematic screening for active TB should be considered in prisons and other penitentiary institutions. (including staff) (disagreement in the group, many wanted a strong recommendation if prevalence >100/100,000)
5. Systematic screening for active TB should be considered in people with untreated fibrotic CXR lesion.
6. In settings where the TB prevalence is ≥100/100,000 in the general population, systematic screening for active TB should be considered among people who are seeking care or who are in care and belong to selected risk groups (see remarks, including staff)
25
26
Risk factor Relative Risk of TB Poor health outcomes related to risk factor
Underweight
(BMI <18.5)
3.2 (95% CI, 3.1–3.3) Increased risk of death and TB relapse; systematic reviews, no pooled
estimate
Gastrectomy or
jejunoileal bypass
No pooled estimate
Gastrectomy: Range 2–5.
Gastric bypass: Range 27–63
Increased risk of death associated with undernutrition (see
”Underweight”)
Diabetes mellitus 3.1 (95% CI, 2.3–4.3) Pooled relative risk of TB treatment failure or death from systematic
review: 1.69 (95% CI, 1.36–2.12) and relapse: 3.89 (95% CI, 2.43– 6.23)
Alcohol dependence 2.9 (95% CI, 1.9–4.6) Higher risk of TB treatment failure and relapse (six studies) and death
during treatment (seven studies); systematic review, no pooled estimate
Tobacco smoking 2.0 (95% CI, 1.6–2.5) Increased risk of death; systematic review, no pooled estimate
Chronic renal failure
or haemodialysis
No pooled estimate; range 10-25 Increased risk of death; systematic review, no pooled estimate
Intravenous drug use No pooled estimate Increased risk of death; systematic review, no pooled estimate
Solid organ
transplantation
No pooled estimate; range, 20–74 No published data
Old age No pooled estimate, prevalence
surveys report increased risk with age
Increased risk of death; systematic review, no pooled estimate
Previously treated TB No pooled estimate Retreatment cases have higher risk of poor outcomes and higher risk of
MDR-TB
Pregnancy No pooled estimate Infants of mothers with TB have increased risks of premature birth and
perinatal death;, pregnant women with TB are more likely to have
complications during pregnancy; initiating TB treatment earlier in the
pregnancy is associated with better maternal and infant outcomes than
late initiation
Conditional recommendations, cont.
7. A. Systematic screening may be considered for geographically defined sub-populations with extremely high levels of undetected TB (>1% prevalence)
B. Systematic screening may be considered also for other sub-populations with very poor health care access, such as urban slum dwellers, homeless people, people living remote areas with poor access, indigenous populations, migrants, and other vulnerable groups.
27
Algorithm options
Algo
rithm
Prevalence = 1% (1,000/100,000)
100,000 persons screened (1,000 true C+ cases)
Screening Diagnosis % of true cases
detected True positive False positive
Positive predictive
value
1
Cough >2-3 w SSM 21% 214 105 67%
GXP 32% 324 52 86%
1. Cough >2-3 w 2. CXR
SSM 19% 193 46 81%
GXP 29% 291 23 93%
2
Any symptom SSM 47% 470 640 42%
GXP 71% 710 320 69%
1. Any symptom 2. CXR
SSM 42% 423 281 60%
GXP 64% 639 141 82%
3
CXR: Any abn. SSM 60% 597 487 55%
GXP 90% 902 244 79%
CXR: TB abn. SSM 53% 529 210 72%
GXP 80% 800 105 88% 28
Number of tests and cost
29
Screening Diagnosis %
detected PPV NPV
CXR per TP
SSM per TP
GXP per TP
Test cost per TP
Cough >2-3 weeks SSM 21% 67% 97.4% 0 26 0 52
GXP 32% 86% 99.5% 0 0 17 173
1. Cough >2-3 w. 2. CXR GXP 29% 93% 99.5% 19 0 9 119
Any symptom GXP 71% 69% 99.8% 0 0 46 461
1. Any symptom 2. CXR GXP 69% 90% 99.1% 48 0 21 286
CXR: TB abnormalities GXP 80% 88% 99.4% 125 0 14 330
Screening Diagnosis %
detected PPV NPV
CXR per TP
SSM per TP
GXP per TP
Test cost per TP
Cough >2-3 weeks SSM 21% 80% 94.9% 0 14 0 28
GXP 32% 93% 98.9% 0 0 9 91
1. Cough >2-3 w. 2. CXR GXP 29% 96% 98.9% 10 0 5 65
Any symptom GXP 71% 82% 99.6% 0 0 23 234
1. Any symptom 2. CXR GXP 69% 95% 98.3% 24 0 11 148
CXR: TB abnormalities GXP 80% 94% 98.7% 62 0 8 169
Prev 1%
Prev 2%
Test cost assumptions
SMS x 2 = 2 $. Digital CXR = 1.5 $. GXP=10 $
Number of tests per true positive case
0
200
400
600
800
1000
1200
1400
1600
1800
2000
0,0% 0,5% 1,0% 1,5%
Nu
mb
er/1
00
,00
0 s
cree
ned
; n
um
ber
nee
ded
to
scr
een
TB prevalence in screened population
TB cases (c+)
True positive
False positive
Number needed to screen G
o a
hea
d !
Screen: Cough >2weeks (35% sens. ; 95% spec.) Diagnosis: SSM (61% sens. ; 98% spec.)
30
500/100,000 C
au
tio
n !
Screen: Cough >2weeks (35% sens. ; 95% spec.) Diagnosis: GXP (92% sens. ; 99% spec.)
31
0
200
400
600
800
1000
1200
1400
1600
1800
2000
0,0% 1,0% 2,0%
Nu
mb
er/1
00
,000
scr
een
ed;
nu
mb
er
ne
ed
ed
to
scr
ee
n
TB prevalence in screened population
TB cases (C+)
True Positive
False Positive
NNS
500/100,000
Ca
uti
on
!
32
0
200
400
600
800
1000
1200
1400
1600
1800
2000
0,0% 1,0% 2,0%
Nu
mb
er/1
00
,00
0 s
cree
ned
; n
um
ber
nee
ded
to
scr
een
TB prevalence in screened population
CXR screen - GXP diagn Screen: CXR Suggestive of TB - sens 0.87 spec 0.89
Conf tests: GXP sens 0.92 spec 0.99
TB cases (C+)
True Positive
False Positive
NNS
Screen: CXR (87% sens. ; 89% spec.) Diagnosis: GXP (92% sens. ; 99% spec.)
500/100,000
Ca
uti
on
!
33
Screen: Cough >2weeks 35% sensitivity 95% specificity
Diagnosis: SSM 61% sensitivity 98% specificity
Algorithms for children
• TB contacts under the age of five years and HIV-infected children: symptom screening (any symptom : cough, weight-loss/failure to
thrive, fever, fatigue) chest radiography may be added to the initial screening.
• Community screening: Children who are TB contacts or HIV-positive follow algorithm above.
Children over the age of 10 years follow adult algorithm
No screening recommended for children <10 who are not contacts or
HIV infected No empirical data on screening test accuracy, but likely to have very low
specificity, since pulmonary TB would be a relatively rare cause of lower respiratory symptoms and signs
Most confirmatory diagnostic tests for TB have lower accuracy in children than in adults, likely resulting in a high number of false-positive cases
34
Key principles
1. Basic conditions before initiating screening Quality diagnosis and treatment (including MDR-TB)
is in place and there is capacity to scale up Baseline assessment: epidemiology, health system,
other screening programmes; legal and human rights framework
Baseline assessment suggests that benefits outweigh harm
Reasonable cost and likely cost-effectiveness compared to other possible interventions
2. Prioritization of risk groups for screening
3. Choosing screening/diagnostic algorithm
4. Ethical issues:
Informed consent; Privacy and confidentiality (when possible) Minimize risks of stigma and discrimination
35
Key principles
5. Optimizes synergies across health/social services Integrate with other screening/health
promotion initiatives Bi-directional screening (HIV, diabetes, etc)
6. M&E, re-assessment, re-prioritization, discontinuation Numbers screened / yield = number needed to
screen Contribution to case notification, by risk group Initial default Treatment outcome
36
Next: Operational guidelines • Findings from systematic review 2,3, and 4, plus field experiences
• Steps before and during screening (checklist)
• Tools for prioritization of risk groups
• Tools for choice of screening and diagnostic tool/algorithm
• Ethical considerations; advice on how to obtain informed consent and disclose
information about results
• M&E tools and forms
37
Thanks!
38
Details on review findings
39
Systematic review 1: impact of screening Question Studies Finding Quality Remarks
Imp
ortan
t ou
tcom
es
Impact on case detection?
5 RCTs 16 cross-sectional
-Screening improves case detection -Larger impact if large groups and large coverage
Very low/low
-Large variation, some studies showed no, small or temporary impact. -Indirectness bias
Impact on time to diagnosis and severity at diagnosis?
Delay: 3 RCTs and 3 cross-sectional Severity: 19 cross sectional
-Shorter reported delay -Smaller proportion smear-positive, smear grade 2/3+ and severe CXR.
Very low/low
-Large variation, some studies showed no or very small impact. -Indirectness bias
Critical o
utco
me
s
Impact on treatment outcomes?
6 comparative cohort analyses 1 case control study on TB mortality (miners)
Same treatment outcome in screened and passively detected Reduced mortality (miners)
Very low -Consistent findings -Length time bias, survival bias
Social/economic consequences?
No study - - -
Impact on TB epidemiology in the community
Community: 1 RCT, 2 before-after Contacts: 2 RCTs Prisons/homeless: 3 Secular trend studies
RCT: no impact Before-after: reduced prev./notification Reduced prev. or incidence Declining secular trend
Moderate Low Low/mod. Very low
-Uncertain impact. One large RCT on community screening showed no impact of moderate intensity screening. -Indirectness bias and confounding
40
GRADE tables: case detection
41
Quality assessment
Summary of findings
Importan
ce
No of
cases/population Effect
Quality No of
studies Design
Limi-
tations
Incon-
sistency
Indirect-
ness
Im-
precision
Inter-
vention Control
Relative
(95%CI) Absolute
1. Impact on case detection
1.1. Cross sectional studies assessing proportion of detected cases through screening vs. all notified cases
Community-based screening, including screening in specific risk populations
11
Cross-
sectional
Serious Serious Not
serious
Not
serious
NA NA NA Proportion of cases
detected through
screening of all TB cases,
range across studies: 6% -
86%
VERY
LOW
IMPORTA
NT
Contact investigation
5 Cross-
sectional2
Serious3 Not
serious
Not
serious
Not
serious
NA NA NA Proportion of cases
detected through
screening of all TB cases,
range across studies: 2% -
9% (19% among children
under 10 in one study)
VERY
LOW
42
Quality assessment Summary of findings
Importance No of cases/population Effect
Quality No of
studies Design Limitations
Inconsistenc
y Indirectness Imprecision
Interventi
on Control
Relative
(95%CI) Absolute
1.2. Community randomized trials
CRITICAL
Health extension workers informed community about TB and identified TB symptomatics and collected sputum samples at health posts once a month
over 20 months. Sputum smear positive notification detection rate
1 (Datiko
2009)
RCT Serious Only one
trial
Serious Not
serious
230/178138
(129/100,00
0)
88/118673
(74/100,000
RR 1.55 Case detection rate 53%
higher in intervention
clusters (95% CI; 40 to 65%)
LOW
Community promoters informed about TB symptoms, sputum collection in monthly outreach clinics in rural areas over 1 year. Sputum smear positive
case notification rate
1 (Shargie
2006)
RCT Serious Only one
trial
Serious16 Serious8 159/1276
07
(125/100,
000)
221/225284
(98/100,000)
RR 1.27
(0.81;1.72
)
27 more cases detected per
100,000 population (from 19
fewer to 72 more)
VERY LOW
Repeat information campaigns about TB to the community, decentralised sputum collection points with easy access, sputum collection in health
camps, over 3 years
1 (Ayles
2012)
RCT Not serious Only one
trial
Serious Serious Cases at 0
vs. 3 years
875/946
(0.92)
Cases at 0 vs. 3
years
2024/2181
(0.93)
1.0 No difference in the change
in case notification between
intervention and control
LOW
Door-to-door screening for symptomatics in poor urban areas
1 (Miller
2010)
During 27d Intervention +60d
RCT Not serious Only one
trial
Serious11 Serious N=11249 19: 934/100000 32: 516/100000
N=12304 16: 604/100,000 41: 493/100,000
Rate Ratio
1.55(1.10;
1.99)
1.05(0.56;1.54)
330/100,000 more cases
(from 60 more to 598 more)
during intervention, no
difference 60 days after
LOW
Infants screened 3 monthly through household visits; suspected TB disease was investigated as inpatient
1 (Moyo
2012)
RCT Serious Only one
trial
Serious Not serious 89/4109
PY
(2166/100
000)
36/4372 PY
(823/100,000)
2.6 (1.8-
4.0)
1,317 more cases detected
per 100,000 population per
year (from 659 more to 2470
more per 100,000)
LOW
Key principles - details
Details
43
1. Prerequisites
1. Quality TB diagnosis, treatment, management and patient support is in place. • Low initial default and treatment interruption; uninterrupted drug supply • Capacity to tailor treatment programme to the needs of the screened population
2. Actions to improve “passive case finding” have been pursued:
• Access to free-of-charge TB services of good quality • Optimization of TB diagnosis in health facilities • Enforced notification of all detected TB cases / PPM
3. Benefit of screening clearly outweighs harm, and at reasonable cost.
• Analysis of potential harm done; legal framework, disease control law; labor laws, immigration laws
• Analysis of individual and/or public health gain in relation to investments.
4. MDR considerations
• Only use diagnostic tests that identify drug resistance if a PMDT programme is in place • Only screen for TB in moderate/high MDR-TB prevalence settings if DST andPMDT is in place
5. Health systems opportunity costs: Sufficient additional resources are available or
can be made available without adverse impact on other key function of the health system
44
2. Prioritization of risk groups 1. Potential benefits for the individual; larger if high risk of delaying diagnosis due
to poor health care access and/or high risk of poor treatment outcomes due to delay, e.g. poor immune status.
2. Potential risks for the individual; screening and diagnostic process (time, inconvenience, cost); negative consequences of correct TB diagnosis (stigma, discrimination); false positive diagnosis.
3. Potential impact on transmission, within and beyond the risk group; congregate settings; in-and out migration. Larger risk group=larger impact.
4. Feasibility and acceptability; reaching; screening; starting and completing treatment. More feasible in well-defined risk groups that are possible to reach in a specific location.
5. The number needed to screen
6. Cost in relation to benefit. NNS; screening and diagnostic algorithm; method of reaching people; direct and indirect costs incurred for individuals.
• NOTE: Benefit/harm judgement, cost-effectiveness and feasibility consideration may differ depending who is the main stakeholder: NTP; MoH; other ministries; NGOs; private sector 45
3. Choice of screening/diagnostic algorithm
1. Estimated true/false positive/negative cases, which depends on; • The sensitivity and specificity of the screening tool(s) • The sensitivity and specificity of the diagnostic tool(s) • The prevalence of undetected TB in the screened population.
2. Consequences of each for each risk group considered;
• True positive: Benefits and risk of correct TB diagnosis • False positive: Wrongly labelled and treated for TB • False negative: Wrongly declared not having TB • True negative: Inconvenience and cost of screening
3. Same considerations for MDR-TB, plus:
• Capacity to deal with persons diagnosed with drug resistance • Need for DST in moderate/high MDR-TB prevalance settings
4. Cost and feasibility
• What screening and diagnostic tests are available or can be easily made available • Costs of tests and number of tests needed for each true positive case detected • Operational logistics and workload, required maintenance etc.
46
4. Ethics and risk minimization
1. "First do no harm"
2. Different ethical obligations when screening compared to "passive case finding"
3. Different target populations have different stakes.
4. User involvement in the design
5. Informed consent and appropriate method to inform about results
6. Assess and address risk of discrimination/stigmatization (e.g. migrants, employees)
7. Minimize pain, discomfort, unnecessary worry, time loss, indirect costs.
8. Minimize risk of stigmatization for the target group as a whole
9. Combine with other screening; Proper management of other conditions than TB
10. Address underlying risk factors and social determinants
47
5. Synergies across health/social services
1. Coordinated and integrated, as appropriate; to optimise effective use of resource to increase relevance and attractiveness of the screening
2. Assess collaboration across different health care disciplines
HIV clinics / VCTC Diabetes clinics, antenatal care clinics, substance abuse clinics General OPD, etc.
3. Explore existing screening platforms;
Non-communicable diseases, childhood malnutrition, malaria, HIV, etc.
Social services for special populations (homeless, refugees, remote areas, slum dwellers, etc)
48
6. M&E, re-assessment, re-prioritization
1. Preparation, monitoring and evaluation plan should be in place with basic indicators • Annual review of active screening programme • Define stopping rules before starting active case screening (doing harm, yield too low)
2. Verify assumptions (yield, NNS, accuracy, acceptability, resource requirement)
3. Continuously assess yield / contribution / impact
4. Re-adjust to changing TB epidemiology and risk group distribution
5. Suggested key indicators for M&E
• Proportion screened of those eligible • Proportion screened positive • Proportion tested of those screened positive • Proportion with TB of those tested • NNS • Initial defaulters and treatment success rate
6. Required R&R
• A log of number of people screened in each risk group • (A register of all suspected TB cases subjected to further diagnostic evaluation) • A column in the laboratory register • A column in the treatment register.
7. Operational research – controlled trials
49
50
Poor individual health Negative public health impact
Primary (for the person with disease)
•Death
High TB death rate
• No cure; delayed cure; sequelae; disability
High prevalence of TB (& of sequelae)
• High cost, stigma, discrimination
High societal cost of TB
Secondary
•TB transmission to contacts
High incidence of infection
Tertiary High incidence of TB disease
Σ individual health = public health
Systematic review 3: Number needed to screen Population (# studies) Weighted Mean NNS (Range)
Low Incidence Moderate Incidence Medium Incidence High Incidence
HIV
HIV clinic/cohort (74) 30 (8-391) 61 (5-316) 13 (2-120*) 10 (3-64)
VCT (5) - 37 (8-120)
Household contacts (89)
54 (5-430)
40 (7-355)
25 (3-568)
17 (2-129)
Prisons (44) 1180 (4-2945) 155 (19-191) 110 (7-2762)
Miners (8) 48 (-) 154 (-) - 36 (21-93)
Within health facilities
Diabetes (6) - 2223 (-) 35 (17-54)
Drug users (8) 158 (108-252) 5 (-) 20 (8-20) -
Elder/Nursing homes (7) 120 (68-137) - 7 (-)
Psychiatric Facilities (3) 1049 (32-1275) 111 (-) -
General Inpatient (4) - - 795 (6-3364)
General Outpatient (14) 758 (42 – 30000) 269 (19-806)
Health Care workers (16) 1613 (30-5550) 506 (25-842*) -
Community
Whole community (98) 3922 (137-30865) 669 (15-5594) 603 (25-4286) 100 (16-6355)
Homeless (18) 133 (22-1778) - - -
Refugees (38) 108 (6-1630) 120 (57-291)
Immigrants (26) 235 (3-1262*) - 1206 (198-6250) - 51
NNS in medium and high incidence settings
52
13 25
110
35 20
111
269
506
603
120
10 17 36
7
795
100
0
100
200
300
400
500
600
700
800
Medium Incidence
High Incidence
1,200
Health facility "Community"
Systematic review 4: Acceptability
53
Risk Groups
Number of
studies
weigted.
mean %
Median
%
Homeless 5 96 75
Mentally Ill 2 95 94
Diabetics 2 94 96
Pregnant women 6 94 90
Drug Dependent 3 93 89
Residents of Urban poor areas 11 91 88
Children Under 5 2 87 86
Migrants 8 85 81
Immigration 14 80 79
PLIHIV 17 78 83
Refugee Camps/IDP 2 74 ?
Elderly- Institutionalized 2 72 83
Prisoners 16 72 86
Miners 6 70 84
Indigenous 9 69 89
Persons testing for HIV 5 69 85
Attendees of Health centers 3 68 57
Health Care Workers 5 59 80
Number of tests and cost
54
Screening Diagnosis %
detected PPV NPV
CXR per TP
SSM per TP
GXP per TP
Test cost per TP
Cough >2-3 weeks SSM 21% 67% 97.4% 0 26 0 52
GXP 32% 86% 99.5% 0 0 17 173
1. Cough >2-3 w. 2. CXR GXP 29% 93% 99.5% 19 0 9 119
Any symptom GXP 71% 69% 99.8% 0 0 46 461
1. Any symptom 2. CXR GXP 69% 90% 99.1% 48 0 21 286
CXR: TB abnormalities GXP 80% 88% 99.4% 125 0 14 330
Screening Diagnosis %
detected PPV NPV
CXR per TP
SSM per TP
GXP per TP
Test cost per TP
Cough >2-3 weeks SSM 21% 80% 94.9% 0 14 0 28
GXP 32% 93% 98.9% 0 0 9 91
1. Cough >2-3 w. 2. CXR GXP 29% 96% 98.9% 10 0 5 65
Any symptom GXP 71% 82% 99.6% 0 0 23 234
1. Any symptom 2. CXR GXP 69% 95% 98.3% 24 0 11 148
CXR: TB abnormalities GXP 80% 94% 98.7% 62 0 8 169
Prev 1%
Prev 2%
Test cost assumptions
SMS x 2 = 2 $. Digital CXR = 1.5 $. GXP=10 $