Tuberculosis cohort analysis in Zimbabwe: The need to strengthen patient follow-up throughout the tuberculosis care cascade

INTRODUCTION: Globally people with tuberculosis (TB) continue to be missed each year. They are either not diagnosed or not reported which indicates possible leakages in the TB care cascade. Zimbabwe is not spared with over 12000 missed cases in 2020. A preliminary review of TB treatment outcomes indicated patient leakages throughout the presumptive cascade and undesirable treatment outcomes in selected cities. Chegutu District had pre-diagnosis and pretreatment losses to follow-up while Mutare City among others had 22.0% of outcomes not evaluated in the second quarter of 2021, and death rates as high as 14% were recorded in Gweru District. The problem persists despite training on data analysis and use. The TB cohorts were analysed to determine the performance of the care cascade and the spatial distribution of treatment outcomes in Zimbabwe. METHODS: Using data from district health information software version 2.3 (DHIS2.3), a secondary data analysis of 2020 drug-sensitive (DS) TB treatment cohorts was conducted. We calculated the percentage of pre-diagnosis, and pre-treatment loss to follow-up (LTFU). For TB treatment outcomes, ‘cured’ and ‘treatment completed’ were categorized as treatment success, while ‘death’, ‘loss to follow-up (LTFU), and ‘not evaluated’ were categorized as undesirable outcomes. Univariate analysis of the data was conducted where frequencies were calculated, and data was presented in graphs for the cascade, treatment success, and undesirable outcomes while tables were created for the description of study participants and data quality. QGIS was used to generate maps showing undesirable treatment outcomes. RESULTS: An analysis of national data found 107583 people were presumed to have TB based on symptomatic screening and or x-ray and 21.4% were LTFU before the specimen was investigated. Of the 84534 that got tested, 10.0% did not receive their results. The treatment initiation rate was 99.1%. Analysis of treatment outcomes done at the provincial level showed that Matabeleland South Province had the lowest treatment success rate of 77.3% and high death rates were recorded in Matabeleland South (30.0%), Masvingo (27.3%), and Matabeleland North (26.1%) provinces. Overall, there were high percentages of not-evaluated treatment outcomes. CONCLUSION: Pre-diagnosis LTFU was high, and high death and loss to follow-up rates were prevalent in provinces with artisanal and small-scale mining (ASM) activities. Unevaluated treatment outcomes were also prevalent and data quality remains a challenge within the national TB control program. We recommended strengthening patient follow-up at all levels within the TB care cascade, strengthening capacity-building for data analysis and use, further analysis to determine factors associated with undesirable outcomes and a study on why LTFU remains high.