Patient experience of general practice and use of emergency hospital services in England: regression analysis of national cross-sectional time series data

Abstract

Background The UK Government has introduced several national policies to improve access to primary care. We examined associations between patient experience of general practice and rates of visits to accident and emergency (A&E) departments and emergency hospital admissions in England. Methods The study included 8,124 general practices between 2011-12 and 2013-14. Outcome measures were annual rates of A&E visits and emergency admissions by general practice population, according to administrative hospital records. Explanatory variables included three patient experience measures from the General Practice Patient Survey: practice-level means of experience of making an appointment, satisfaction with opening hours, and overall experience (on 0-100 scales). The main analysis used random-effects Poisson regression for cross-sectional time series. Five sensitivity analyses examined changes in model specification. Results Mean practice-level rates of A&E visits and emergency admissions increased from 2011-12 to 2013-14 (310.3 to 324.4 and 98.8 to 102.9 per 1,000 patients). Each patient experience measure decreased; for example, mean satisfaction with opening hours was 79.4 in 2011-12 and 76.6 in 2013-14. In the adjusted regression analysis, a standard deviation increase in experience of making appointments (equal to nine points) predicted decreases of 1.8% (95% CI: -2.4% to -1.2%) in A&E visit rates and 1.4% (95% CI: -1.9% to -0.9%) in admission rates. This equalled 301,174 fewer A&E visits and 74,610 fewer admissions nationally per year. Satisfaction with opening hours and overall experience were not consistently associated with either outcome measure across the main and sensitivity analyses. Conclusions Associations between patient experience of general practice and use of emergency hospital services were small or inconsistent. In England, realistic short-term improvements in patient experience of general practice may only have modest effects on A&E visits and emergency admissions.