Tuberculosis outcomes during the COVID-19 pandemic in two high-burden countries

The COVID-19 pandemic has disrupted tuberculosis (TB) services globally. Measuring the extent of these disruptions is critical to mitigate their impact and provide valuable information for TB program improvement, particularly in settings hard-hit by both COVID-19 and TB, such as Peru and India. The overall aim of this manuscript-based thesis was therefore to evaluate TB care and outcomes during the COVID-19 pandemic in Peru and India and compare them to pre-pandemic trends. In Manuscript 1, we introduced TB care cascades and outlined data gaps for constructing them. Care cascades are a representation of the proportion of people reaching key milestones in TB care and are a valuable tool, particularly in the context of COVID-19, for measuring TB care delivery disruptions across the care pathway. This provides important context for Manuscript 2, in which we constructed the drug-susceptible (DS) and drug-resistant (DR)TB care cascades prior to and during COVID-19 in Lima province, Peru. We also collected data from patient registries of n=1,513 people receiving TB care in a district of Lima to identify predictors of cascade completion. The final gap in the cascade (those initiating vs. those completing treatment) was the largest gap across all cascades except for in the group receiving TB care entirely during COVID-19, in which the pre-diagnosis gap (estimated burden vs. number diagnosed) was the largest for both the DS-TB and DR-TB cascades. Receiving TB care entirely during COVID-19 was associated with a lower likelihood of cascade completion (OR:0.51, 95%CI:0.29-0.88).Given that COVID-19-related setbacks have made it even more urgent to focus TB elimination efforts on key populations, in Manuscript 3, we further investigated the effect of COVID-19 disruptions on TB specifically in incarcerated populations in Peru, who face a particularly high burden of TB, and who may have experienced COVID-19 disruptions differently than the non-incarcerated population. Using Peruvian National TB Program (NTP) data we conducted interrupted time series analyses in the non-incarcerated, incarcerated and overall population to assess pre- and during COVID-19 trends in TB case notifications compared to their counterfactual values (in the absence of COVID-19). The start of COVID-19 was defined as 16 Mar 2020 (declaration of the national emergency due to COVID-19 in Peru). Our findings indicate a 61.2% (95%CI: 59.9-62.7%) drop in TB case notifications in the first week of the COVID-19 pandemic in the non-incarcerated population and a 17.7% (95%CI: 17.5-17.9%) drop in the incarcerated population. Finally, in Manuscript 4, we assessed in-treatment and post-treatment case-fatality among people diagnosed with TB and enrolled in a private sector TB program in India in the first year of the COVID-19 pandemic (Mar 2020-Mar 2021) and compared this to pre-pandemic estimates among people enrolled in the same program. We collected follow-up data through phone surveys of n=4,000 individuals, of which n=2,962 (74.1%) responded to the survey. We used inverse probability of selection (IPS) weighting to correct for selection bias. We estimated an IPS weighted in-treatment case fatality ratio (CFR) of 6.07% (95%CI: 5.22-6.93%), significantly lower (p=0.036) than the pre-COVID-19 CFR (7.27%, 95%CI: 5.97-8.49%).In conclusion, large pre-diagnosis gaps in the during-COVID-19 TB care cascades in Peru highlight that efforts to increase case detection are urgently needed to regain lost progress on TB elimination. Furthermore, TB burden is extremely high in Peruvian prisons, but TB case notifications were less impacted by COVID-19 disruptions than the general population. Lastly, the observed TB case-fatality in an Indian private sector cohort treated during COVID-19 is above the level needed to reach the 2025 and 2030 End TB Strategy targets for reductions in TB deaths, further highlighting the need to get on track to TB elimination