Yiu-Shing Lau

Associations between Extending Access to Primary Care and Emergency Department Visits: A Difference-In-Differences Analysis

Background Health services across the world increasingly face pressures on the use of expensive hospital services. Better organisation and delivery of primary care has the potential to manage demand and reduce costs for hospital services, but routine primary care services are not open during evenings and weekends. Extended access (evening and weekend opening) is hypothesized to reduce pressure on hospital services from emergency department visits. However, the existing evidence-base is weak, largely focused on emergency out-of-hours services, and analysed using a before-and after-methodology without effective comparators. Methods and Findings Throughout 2014, 56 primary care practices (346,024 patients) in Greater Manchester, England, offered 7-day extended access, compared with 469 primary care practices (2,596,330 patients) providing routine access. Extended access included evening and weekend opening and served both urgent and routine appointments. To assess the effects of extended primary care access on hospital services, we apply a difference-in-differences analysis using hospital administrative data from 2011 to 2014. Propensity score matching techniques were used to match practices without extended access to practices with extended access. Differences in the change in “minor” patient-initiated emergency department visits per 1,000 population were compared between practices with and without extended access. Populations registered to primary care practices with extended access demonstrated a 26.4% relative reduction (compared to practices without extended access) in patient-initiated emergency department visits for “minor” problems (95% CI -38.6% to -14.2%, absolute difference: -10,933 per year, 95% CI -15,995 to -5,866), and a 26.6% (95% CI -39.2% to -14.1%) relative reduction in costs of patient-initiated visits to emergency departments for minor problems (absolute difference: -£767,976, -£1,130,767 to -£405,184). There was an insignificant relative reduction of 3.1% in total emergency department visits (95% CI -6.4% to 0.2%). Our results were robust to several sensitivity checks. A lack of detailed cost reporting of the running costs of extended access and an inability to capture health outcomes and other health service impacts constrain the study from assessing the full cost-effectiveness of extended access to primary care. Conclusions The study found that extending access was associated with a reduction in emergency department visits in the first 12 months. The results of the research have already informed the decision by National Health Service England to extend primary care access across Greater Manchester from 2016. However, further evidence is needed to understand whether extending primary care access is cost-effective and sustainable.