TB epidemiological reviews: baseline to setting country-level targets 9 th Technical Advisory Group and National TB Programme Managers' Meeting Manila,

TB epidemiological reviews: baseline to setting country-level targets

9th Technical Advisory Group and National TB Programme Managers' Meeting

Manila, 10 December 2014

Charalampos (Babis) Sismanidis Philippe Glaziou

GLOBAL TB PROGRAMME

WHO Global Task Force on TB Impact Measurement

www.who.int/tb/advisory_bodies/impact_measurement_taskforce

National TB Programmes of many countries& key technical and funding agencies

What is the aim of a TB epidemiological review?

• To offer a country-level, systematic and standardised assessment of:– The strengths & gaps of the quality & coverage of the TB

surveillance system as well as direct measurement of disease burden

– The best estimates of the level of, and trends in, TB disease burden– A plausible interpretation of how TB disease burden is influenced

by prevention and treatment interventions implemented by NTP and partners?

• To inform programme reviews, "epidemiological stage" of Concept Note submission to the Global Fund and target setting (at least short-, but also useful for longer-term)

What is the aim of a TB epi review? (cont.)

• Standardised terms of reference (available since early 2013)

• Four objectives, with suggested analytical tasks per objective:

1. Assessing quality & coverage of the surveillance system

2. Understanding the level of, and trends in, TB burden3. Plausible interpretation of how various factors drives

TB burden (control activities and risk factors)

4. Development of investment plan to address gaps identified

• Used in about 21 countries globally so far

Objective 1 assessing quality and coverage of surveillance

Describe and assess current national TB surveillance and vital registration systems, with particular attention to their capacity to measure the level of and trends in TB disease burden (incidence and mortality).

http://www.who.int/tb/publications/standardsand-benchmarks/en/

Summary results: 21 countries (Jan 13-Jul 14)

Systematically assess quality of paper-based TB data

(Service Availability & Readiness Assessment)

Move from paper to electronic case-based system

Inventory study to measure under-reporting, mandatory

reporting of TB, PPM expansion

Advocate for use of VR systemImprove surveillance data for

burden in MDR-TB, TB/HIV, children

Where has the checklist been used?

3 WPRO countries: Fiji, the Philippines, Viet Nam

Objective 2 understanding the level of, & trends in, TB burden

Assess the level of, and trends in, TB disease burden (incidence, prevalence, mortality) through systematic collation and compilation of available surveillance, survey, programmatic and other data.

http://www.who.int/tb/publications/

understanding_and_using_tb_data/en/

Do case notification rates in China reflect incidence?

TB notifications

Evidence that prevalence declined in China (1990 – 2010)

Wang, L. et al. (2014). Lancet, 383(9934), 2057–64.

Notifications // incidence? … probably not

TB notifications

TB Mortality3rd national survey

DSP

Evidence of TB under-reporting in China before the SARS epidemic

Reporting reform following SARS epidemic

Notification rate

Estimated incidence rate

TB Mortality

Time trends in case notifications (national)do trends in notification reflect trends in incidence?

Case notification rates per 100,000 population (black: new, all forms; blue: new, bacteriologically-confirmed; red: new, cl inically-diagnosed; green: new, extra-pulmonary). Data source: WHO TB database

Contribution of active case finding (e.g. contact tracing), PPM over time?

DOTS expansion

Quantifying gap between incidence and case notification:

under-reporting in India

46% of cases on treatment not known to NTP

PloS One, 2011; 6(9): e24160.

Objectives1. To explain and promote the role and value of inventory studies to TB

care and control2. To explain (i) major alternative study design & (ii) key issues

concerning the implementation and analysis of inventory studies3. To facilitate the development of a draft protocol outline for a TB

inventory study

Design and protocol development workshop: TB inventory studies to measure under-reporting

of TB cases

24-26 September, 2014

Getting to the non-NTP & closing gap in surveillance

Under-reporting: (199+99+9)/1980 = 16%

An estimated additional cases 473 (394 – 565)

First protocol development workshop for TB inventory studies (Bali, September 2014): summary of key design aspects, timelines and

requirements

China Indonesia Pakistan Philippines Thailand Viet Nam

Objectives Under-reporting (retrospective)

Under-reporting(prospective)

Under-reporting(prospective)

Data quality assessment (NTP and non-NTP)

Under-reporting (retrospective & continuous)

Under-reporting & incidence (prospective)

Case definitions

Bact-conf (all ages)

• All-form TB

• Bact-conf• Children

Children • All-form TB

• Bact-conf• Children

• All-form TB

• Bact-conf• Children

• All-form TB

• Bact-conf• Children

Timelines• Protocol• Study end

• Q1 2015• Q4 2015

• Q4 2014• Q4 2015

• Q4 2014• Q4 2015

• Q2 2015• Q4 2016

• Q4 2014• Q4 2015

• Q4 2014• Q4 2015

Support required

• TA • TA• Funding

• TA• Funding

• TA• Funding

• TA• Funding

• TA• Funding

Time trends in case notifications (sub-national)geographical heterogeneity or diagnostic capacity?

Time series of TB case notification rates per 100,000 (blue: all forms, red: bacteriologically-confirmed), by division/state, 2000-2013. Data source: NTP database

Age-specific time trends in case notificationswho are we notifying?

Time series in new, bacteriologically-confirmed, age-specific case notification rates per 100,000, 1995-2012. Data source: WHO TB database

Age- & sex-specific trends in TB prevalence rates identifying higher at risk groups

Ratio of prevalence to notification rates identifying case detection differentials

Smear-positive TB prevalence rate per 100,000 to annual (2012) smear-positive TB case notification rate per 100,000 ratios; overall and broken down by age and sex. Data source: 1st National TB Prevalence Survey, Myanmar

all

15-24

25-34

35-44

45-54

55-64

65+

male

female

1

2

3

4

Pre

vale

nce

to

no

tific

atio

n r

atio

Population

TB surveillance datainternal consistency

TB case notification ratios of: (i) new pulmonary to extra-pulmonary, (ii) new male to female, and (i ii) new to previously treated, 2009-2013. Data source: WHO TB database

Year Pulmonary:

Extra-pulmonary1 Male:Female2 New:Retreated3

2009 2.9 2.0 13.2 2010 3.5 1.9 12.8 2011 3.8 1.9 12.4 2012 5.6 1.9 12.2 2013 6.7 NA 11.1

1 Ratio of new pulmonary (bacteriologically-confirmed and clinically-diagnosed) to extra-pulmonary notified TB cases; 2 Ratio is calculated only among new bacteriologically-confirmed case notifications, since age and sex disaggregated data are not available for those clinically-diagnosed and extra-pulmonary; 3 Ratio of total new and relapse to those previously treated (excluding relapses); NA: data not available due to the transition of the recording and reporting system in the country according to the 2013 WHO revision. Data source: WHO TB database

Objective 3 How do determinants influence TB burden?

Assess whether recent trends in TB disease burden indicators are plausibly related to changes in TB-specific interventions taking into account external factors including economic or demographic trends.

Which are the upward and which the downward drivers?

TB determinants HIV disease burden

TB determinants economic growth

Trends in GDP per capita in Nigeria, 1960-2010. Data source: World Bank

TB determinants inequalities in health service provision

Maternal health service coverage by wealth quintile in Nigeria, 2012. Data source: WHO GHO, Nigeria: Equity Profile

TB determinants proxies for performance of health system

Trends in under-5 mortality Nigeria, 1960-2010. Data source: World Bank

TB determinants proxies for performance of health system (cont.)

Scatterplot of under-5 mortality rate per 1,000 l ive births against GDP per capita (2010). Each blue dot represents a country pair of data points. Nigeria is shown in red. The rectangle encloses countries with GDP between 1,000 and 1,500 USD per capita. Data source: World Bank

TB determinants age and sex structure of general population

Population pyramids of the general population representing the distribution of the population by age and sex. Data source: UNPD database

Putting it all together programmatic implications (& target definition)

Key finding Programmatic implication

High prevalence to notification rate ratios TB diagnostic and treatment services not

available in many primary health care settings PPM activities not yet reaching their full potential

Improve case finding overall through: decentralisation of TB services; continued expansion of PPM and

making reporting of TB mandatory. Zonal and state level variability in:

case notification rates treatment outcomes

Identify zones and states in order to: take corrective action where

required; learn lessons from good performers

and implement on the rest. Higher disease burden in certain subgroups:

Men (prevalence survey) Younger age groups (prevalence survey) Low case notification rates are observed in

children considering the high disease burden of their parents (the young to middle age groups found in the prevalence survey)

Target intensified case finding activities to these subgroups.

Disease burden due to TBtime trends in incidence

Trends in estimated TB incidence rates per 100,000 population (blue). The blue ribbon shows the uncertainty range. The black line shows case notification rates (all forms). Data source: WHO TB database

Disease burden due to TBtime trends in prevalence

Trends in TB prevalence rates per 100,000 population (black l ine). The red ribbon denotes the uncertainty around TB prevalence estimates. The horizontal dashed l ine denotes the Stop TB Partnership target of a 50% reduction in prevalence rates by 2015 compared with 1990. Data source: WHO TB database

Disease burden due to TBtime trends in mortality

Trends in TB mortality (excluding HIV) rates per 100,000 population (black l ine). The red ribbon denotes the uncertainty around TB mortality estimates. The horizontal dashed l ine denotes the Stop TB Partnership target of a 50% reduction in mortality rate by 2015 compared with 1990. Data source: WHO TB database

Objective 4 investment plan: the example of Indonesia

Define the investments needed and define associated targets to: (i) strengthening surveillance and (ii) directly measuring trends in TB disease burden

Objective 4DRAFT investment plan: the example of MyanmarImprove case detection• Expand PPM with private hospitals, specialists (e.g. paediatricians) and

pharmacies (target: xx% of private hospitals linked with NTP by xxxx)

Improve M&E capacity and data quality• Enhancing capacity on good data management and analytical practices (target: xx

number of epidemiology and statistics workshops/courses, xx number of staff in M&E team)

• Advocate for the transition from a paper to an electronic case-based recording and reporting system

(target: electronic case-based pilot by xxxx)

Improve direct measurement of disease burden• Supporting the development of a high-quality national vital registration system

with standard coding of cause of death (target: liaison officer to report on TB deaths from VR, xx number of staff trained on ICD-10)• Conduct studies to improve direct measurement of TB disease: repeat TB

prevalence survey in 2016/2017 (target: repeat prevalence survey by xxxx)

Conclusion & next steps• Epidemiological and impact reviews offer a unique

opportunity to: – conduct a baseline assessment of the strengths and weaknesses of the

surveillance system (repeat review in some years t monitor progress);– understand, use & improve the quality of TB, & other relevant, data; – identify data gaps for direct measurement of TB burden;– set targets on improving: (i) quality and coverage of surveillance,

and (ii) direct measurement of disease burden.

• Next steps:– Discussion with technical partners on revision of ToR's of

epidemiological reviews (e.g. include projections on key indicators)– Enhance global and regional capacity to conduct epidemiological

reviews (e.g. model examples, train a small roster of consultants)– Promote the wide use of handbook on "understanding and using TB

data" (e.g. workshops)

Acknowledgements

• National TB Control Programmes (particularly Indonesia, Myanmar & Nigeria)

• Katherine Floyd

• Ikushi Onozaki

• Irwin Law

• Hazim Timimi

• Emily Bloss

• Suzanne Verver

• Eveline Klinkenberg

• Norio Yamada

• Chikwe Ihekweazu

• Ananta Nanoo

• Global Fund

• USAID

• UNITAID

• TB CARE