Factors associated with low tuberculosis case notification and treatment success at health facilities of Zambia: a cross-sectional study

Abstract

Introduction: Early detection and successful treatment of people with Tuberculosis (TB) prevents millions of deaths globally. Yet, gaps persist in the detection and treatment of TB. Zambia’s weak health system only exacerbates matters with the country having one of the highest TB burdens in the world. Furthermore, 58 per cent of the identified TB patients in Zambia are co-infected with HIV making it a double public health burden. Aim: To determine health systems factors associated with low tuberculosis case notification and treatment success in health facilities of Zambia as well as determine their variations in the effect size of associations. Methods: The study used secondary health facility data from the 2019 Health Facility Listing Survey and 2017 and 2018 Health Management Information System data sets. A cross-sectional design was used to analyze data from 81health facilities from 9 provinces of Zambia. Data was managed using STATA version 14. Linear regression analysis was used to analyze factors associated with low TB case notification and treatment success while quantile regression and principal component analysis were used to determine the effect size of these associations. Results: Linear regression analysis indicated that low TB case notification was positively associated with personnel(P-value 0.00, CI =0.120.62), negatively associated with rural clinic ( p-value 0.00, CI= -2.91, 0.84),negatively associated with 3rd level hospital(p-value 0.05, -5.12, 0.15) and negatively associated with having no TB clinic(p-value 0.01, CI= -1.57,-2.48) while low Treatment Success was positively associated with personnel(p-value 0.02, CI= 0.06, 0.65)and population(p-value 0.00, CI=0.13, 0.58) but negatively rural clinic(p-value 0.00, CI= -3.07,-1.55) and having no TB clinic(p-value 0.00, CI= -2.10,-0.03). When analyzed by year, both low TB notification and Treatment success were associated with personnel in 2017 and 2018 respectively. While results from quantile regression showed that for facilities at the 25th percentile of case notification or treatment success, having an addition staff was associated with 3 times year increase in notification or treatment success than facilities at all other quartiles. Similarly, results from simple regression using principal component analysis showed that those facilities that had equipment at the 3rd quarter were 2 and 3 times higher to notify TB and treat it compared to those facilities that had equipment at the lower quarters, however, when the equipment variable was run in a multiple regression, results show that it was insignificant. iv Conclusion: Low TB case notification and treatment success still remains a challenge in health facilities of Zambia. Using systems thinking approach is thus cardinal in understanding and tackling health systems barriers affecting TB control programs.