Impact of Community-Wide Tuberculosis Active Case Finding and Human Immunodeficiency Virus Testing on Tuberculosis Trends in Malawi

Abstract

Background Tuberculosis case-finding interventions are critical to meeting World Health Organization End TB strategy goals. We investigated the impact of community-wide tuberculosis active case-finding (ACF) alongside scale-up of HIV testing and care on trends in adult tuberculosis case notification rates (CNRs) in Blantyre, Malawi. Methods Five rounds of ACF for tuberculosis (1-2 weeks of leafleting, door-to-door enquiry for cough and sputum microscopy) were delivered to neighbourhoods (“ACF areas”) in North-West Blantyre between April 2011 and August 2014. Many of these neighborhoods also had concurrent HIV testing interventions. The remaining neighbourhoods in Blantyre City (“non-ACF areas”) provided a non-randomised comparator. We analyzed TB CNRs from January 2009 until December 2018. We used interrupted time series analysis to compare tuberculosis CNRs before ACF and after ACF, and between ACF and non-ACF areas. Findings Tuberculosis CNRs increased in Blantyre concurrently with start of ACF for tuberculosis in both ACF and non-ACF areas, with a larger magnitude in ACF areas. Compared to a counterfactual where pre-ACF CNR trends continued during ACF period, we estimated there were an additional 101 (95% confidence interval [CI] 42 to 160) microbiologically confirmed (Bac+) tuberculosis diagnoses per 100,000 person-years in the ACF areas in three and a half years of ACF. Compared to a counterfactual where trends in ACF area were the same as trends in non-ACF areas, we estimated an additional 63 (95% CI 38 to 90) Bac+ diagnoses per 100,000 person-years in the same period. Interpretation Tuberculosis ACF was associated with a rapid increase in people diagnosed with tuberculosis in Blantyre.

Community-based active-case finding for tuberculosis: navigating a complex minefield

“…TB REACH projects are also preferentially focused on marginalised communities with poor routine health access [34]. Moreover, funding was of short duration and nearly all studies only investigated trends in case notification rates during the implementation of the programme (and occasionally before) [22]; to the best of our knowledge, only two studies (one done in Brazil [37] and one in Malawi [38]) compared trends in the post-ACF period. This is important as historical evidence and modelling show us that the effects of once-off ACF on underlying TB epidemiology and diagnosis are only likely to be temporary, with repeated rounds and multiyear follow-up required for lasting effect on underlying population epidemiology [39,40].…”

“…2 Interpretation of trends in tuberculosis case notification rates following implementation of active case-finding programmes. Note: The figure is illustrative, but based on data from Blantyre, Malawi, analysed by Burke et al [38]. ACF active case finding Mtb has at best only moderate predictive precision for identifying individuals who will progress to development of TB [79], at the population level, distributions of positive tuberculin skin tests, IGRAs, or newer tests can give an "epidemiological signal".…”