INnovative data Science to Impact the TB Epidemic - INSITE

PROJECT SUMMARY There are limited opportunities in Africa to work with routine clinical and health data at scale, and to develop the expertise to process these data in an environment in which both the burden of disease and the approach to enumeration and coding are very different to other settings where routine data are exploited for epidemiological

analyses. Tuberculosis disease (TB) is the most important HIV-associated comorbidity in South Africa and the leading cause of death in people living with HIV (PLH). Maternal TB (pregnancy and postpartum period) deleteriously affects morbidity and mortality in women and infants. Pregnancy outcomes are worse, exacerbated

by HIV co-infection which is present in 60%, and there is increased risk of perinatal TB and HIV transmission. Rifampicin-resistant (RR)-TB with/without fluoroquinolone resistance is a growing global public health concern with prevalence rates at 4.2% of TB cases in the Western Cape Province, South Africa. South Africa was prompt

to include novel and repurposed second-line therapeutic agents against RR-TB in routine guidelines, including pregnant and lactating people. The safety of these agents during pregnancy and lactation is uncertain but the rising incidence of RR-TB has seen the use of these agents during pregnancy at scale. The concomitant use of

antiretroviral therapy (ART) for HIV treatment raises concerns about potential drug-drug interactions with safety and efficacy implications. Maternal TB remains under-prioritized globally despite the burden of HIV and TB in those of child-bearing age. In this context of high infectious burden in which novel/repurposed anti-mycobacterial agents for the prevention

and treatment of TB are used at scale during pregnancy, often co-administered with ART, the Western Cape is uniquely positioned in hosting the Provincial Health Data Centre (PHDC), the first comprehensive linked health information exchange in Africa. The PHDC leverages the use of a province-wide unique patient identifier issued

to users of the health services to harmonise all electronic health data from routine health information systems in public sector facilities, including laboratory, pharmacy, administrative and other clinical data. These data are integrated and curated to enable inference of health conditions using phenotype algorithms, as well as to enable

the consolidation of co-morbidities and construction of patient-level cascades that can be used to monitor and analyse health service utilisation patterns and healthcare outcomes. The partnership between the Centre for Infectious Disease Epidemiology & Research, University of Cape Town and the Health Intelligence Directorate of the Western Cape Government which houses the PHDC offers a

unique opportunity to develop and evaluate a robust maternal TB data platform, using data science to i) create actionable clinical tools to optimize person-level interventions and monitor programs; ii) generate large, linked cohorts to address the epidemiological questions of maternal TB and assess the impact of policy interventions

and clinical tools at the population and individual levels.