Absenteeism of health workers in developing countries is common and can severely undermine the reliability of the health system. Therefore, it is important to understand where the prevalence of absenteeism is high. We develop a simple imputation method that combines a Service Delivery Indicators survey and a Service Provision Assessment survey to estimate the prevalence of absenteeism of health workers at the level of regions in Tanzania. The resulting estimates allow one to identify the regions in which the prevalence of absenteeism is significantly higher or lower than the national average and help policymakers determine priority areas for intervention.
Keywords: imputation; primary health facility; random-effects probit; service delivery indicator; sub-Saharan Africa.
© 2018 John Wiley & Sons, Ltd.