More Doctors, better health? Considering doctor numbers in the Mais Medicos Programme

Abstract

Worldwide, there are insufficient primary care physicians to provide accessible, high-quality primary care services. The evidence base on how effective primary care can influence population health outcomes is also weak. We estimate the effect of an increasing in primary care physicians using the Programa Mais Médicos. Although previous studies have used the Programa Mais Médicos to analyse the impact of a primary care physician supply using a binary treatment indicator, this mask substantial variation in the numbers of Mais Médicos physicians that regions received. Therefore, we exploit the variation in physicians allocated to each Brazilian microregion to identify the impact of Mais Médicos primary care physicians, as a continuous treatment effect, on hospitalisations and mortality rates (per 100,000 population), and ambulatory care sensitive hospitalisation and mortality rates. Our estimates can then be considered a dose-response unlike previous work, and the methodological approach also allows us to avoid some limitations of previous work. We do so by using a panel dataset in which we observe monthly observations of all Brazilian microregion over the period 2008-2017. The variation in Mais Médicos physicians across microregions is exploited using the generalised synthetic control estimator, which has benefits over difference-in-difference estimators. The generalised synthetic control methodology estimates suggest a limited effect of primary care physicians impacting health outcomes. A separate estimate of the impact of Mais Médicos physicians are estimated for each state, and we find that in each Brazilian state, Mais Médicos has no statistically significant impact on hopsitalisations or mortality rates, either total or ambulatory care sensitive conditions. However, descriptive evidence might suggest that regions with low numbers of primary care physicians pre- Programa Mais Médicos roll-out may have benefitted more from receiving Mais Médicos physicians, and this supports findings of previous work. There are several explanations of these results including: substitution of other health professionals, impacts materialising over the longer-term, rather than the short-term effects estimated here, primary care physicians having a limited influence over health outcomes, and also poor allocation of Mais Médicos physicians.