Are more GPs associated with a reduction in emergency hospital admissions?

Abstract

Background: Recent studies have found an association between access to primary care and accident and emergency attendances, with better access associated with fewer attendances. Analyses of an association with emergency admissions however have produced conflicting findings. Aim: We investigate whether emergency admission rates in an area are associated with (i) the number of GPs, and (ii) mean size of GP practices. Design and Setting: Analysis was conducted utilising Hospital Episode Statistics, the numbers of GPs and GP practices, ONS population data, Quality and Outcomes Framework (QoF) prevalence data, and Index of Multiple Derivation data, from 2004/5 to 2011/12, for all practices in England. Method: Regression analysis of panel data with fixed effects to address (i) a potential two-way relationship between the numbers of GPs and emergency admissions, and (ii) unobservable characteristics of GP practices. Results: There is not a statistically significant relationship between the number of GPs in a local area and the number of emergency admissions when analysing all areas. However, in deprived areas, a higher number of GPs is associated with lower emergency admissions. There is also a lower emergency admission rate in areas in which practices are on average larger, holding constant GP supply. Conclusions In deprived areas an increase in GPs was found to reduce emergency admissions but does not do so elsewhere. Areas in which GPs became concentrated into larger practices experienced reduced levels of emergency admissions, all else equal.