Impact of user fees on maternal health service utilization and related health outcomes: a systematic review

Objective To assess the evidence of the impact of user fees on maternal health service utilization and related health outcomes in low- and middle-income countries, as well as their impact on inequalities in these outcomes. Methods Studies were identified by modifying a search strategy from a related systematic review. Primary studies of any design were included if they reported the effect of fee changes on maternal health service utilization, related health outcomes and inequalities in these outcomes. For each study, data were systematically extracted and a quality assessment conducted. Due to the heterogeneity of study methods, results were examined narratively. Findings Twenty studies were included. Designs and analytic approaches comprised: two interrupted time series, eight repeated cross-sectional, nine before-and-after without comparison groups and one before-and-after in three groups. Overall, the quality of studies was poor. Few studies addressed potential sources of bias, such as secular trends over time, and even basic tests of statistical significance were often not reported. Consistency in the direction of effects provided some evidence of an increase in facility delivery in particular after fees were removed, as well as possible increases in the number of managed delivery complications. There was little evidence of the effect on health outcomes or inequality in accessing care and, where available, the direction of effect varied. Conclusion Despite the global momentum to abolish user fees for maternal and child health services, robust evidence quantifying impact remains scant. Improved methods for evaluating and reporting on these interventions are recommended, including better descriptions of the interventions and context, looking at a range of outcome measures, and adopting robust analytical methods that allow for adjustment of underlying and seasonal trends, reporting immediate as well as longer-term (e.g. at 6 months and 1 year) effects and using comparison groups where possible.