Contributions of a community health worker program in improving district health system performance in South Africa, using routine health service variables in a quasi-experimental study design

Abstract

BACKGROUND : Globally low- and middle-income countries are investing in community health worker (CHW) programs. These programs are viewed as fundamental to ensuring universal health coverage for all. Where CHW programs are doing what they should be doing and doing it adequately, they should make a difference in population health outcomes, reflecting health system changes. However, there are not many studies exploring the contributions and effectiveness of large-scale, comprehensive community health worker programs on health system performance. METHODS : This study takes place in the Ekurhuleni health district, in urban South Africa. A before (pre) and after (post) intervention study (quasi-experimental) using routine health service variables (secondary data) from district clinics over nine years was conducted. Performance of intervention clinics with community health worker teams that had more than 60% population coverage were compared to similar control clinics that had no teams or teams with less than 30% coverage. RESULTS : Both groups of clinics generally improved over time for routine health service variables for maternal, child health, infectious diseases, and cervical cancer programmes. Over nine years, intervention clinics showed more improved performance, statistically significant, in six health variables while controls improved in four. But importantly, intervention clinics improved proportionately better pre to post in seven of the nine (78%) variables studied compared to controls, demonstrating reduced diarrhoeal disease and pneumonia in children, better Vitamin A coverage, fewer severely malnourished children, better testing for HIV, and better screening for Tuberculosis and cervical cancer. Illustrating the comprehensive range of services provided by CHWs, the seven routine health service variables represented maternal, child health, infectious and non-communicable diseases. We also observed less uncontrolled hypertensive and diabetics in intervention clinics compared to control clinics. CONCLUSION : Large-scale, sustained and comprehensive community health worker programs adequately covering populations contributed to improved urban district health system performance in Ekurhuleni, demonstrating their effectiveness. This was explored through improvements in health system performance over time. These findings, using routine health service variables, have policy implications for financial and other resource allocations in health districts in low- and middle-income countries.