Factors associated with community health worker performance differ by task in a multi-Tasked setting in rural Zimbabwe

Background: Zimbabwe, like most low-income countries, faces health worker shortages. Community health workers (CHWs) bridge this gap by delivering essential health services and nutrition interventions to communities. However, as workloads increase, CHWs' ability to provide quality services may be compromised.We studied influences upon CHWs' performance related to pregnancy surveillance and nutrition and hygiene education in rural Zimbabwe. Methods: In the context of a cluster-randomized trial conducted in 2 rural districts between November 2012 and March 2015, 342 government-employed CHWs identified and referred pregnant women for early antenatal care and delivered household-level behavior change lessons about infant feeding and hygiene to more than 5,000 women. In 2013, we conducted a survey among 322 of the CHWs to assess the association between demographic and work characteristics and task performance. Exploratory factor analyses of the Likert-Type survey questions produced 8 distinct and reliable constructs of job satisfaction and motivation, supervision, peer support, and feedback (Cronbach a range, 0.68 to 0.92). Pregnancy surveillance performance was assessed from pregnancy referrals, and nutrition and hygiene education performance was assessed by taking the average summative score (range, 5 to 30) of lesson delivery observations completed by a nurse supervisor using a 6-item Likert-Type checklist. Poisson and multiple linear regressions were used to test associations between CHW demographic and work characteristics and performance. Results: CHWs who referred more pregnant women were female, unmarried, under 40 years old, from larger households, and of longer tenure. They also perceived work resources to be adequate and received positive feedback from supervisors and the community, but they were less satisfied with remuneration. CHWs with high scores on behavior change lesson delivery were from smaller households, and they received more supportive supervision but less operational supervision. Measures of job satisfaction and motivation were not associated with either task. Conclusion: Among CHWs responsible for multiple tasks in rural Zimbabwe, factors associated with performance of one task were not the same as those associated with performance of another task. Our methods and findings illustrate ways to examine heterogeneity in CHW performance and to identify organizational factors associated with quality of program delivery.