Background:
In 2002 British Columbia, Canada began redistributing its hospital services.
Objective and Design:
We used administrative data and interrupted time series analyses to determine how recent hospital closures affected patient outcomes.
Subjects:
All adult acute myocardial infarction (AMI), stroke, and trauma events in British Columbia between fiscal years 1999 and 2013. Cases were patients whose closest hospital closed. Controls were matched by condition, year of event, and condition-specific hospital volume where treatment was received.
Measures:
Thirty-day mortality and hospital bypass rates.
Results:
We matched 3267 AMI, 2852 stroke, and 6318 trauma cases to 1996, 1604, and 3640 controls, respectively. The 30-day mortality rate at baseline was 7.0% [95% confidence interval (CI), 4.0%–10.1%] for AMI, 5.3% (95% CI, 2.4%–8.1%) for stroke, and 1.2% (95% CI, 0.3%–2.1%) for trauma controls. The 30-day mortality rate for cases was 14.3% (95% CI, 7.1%–21.7%) for AMI, 12.0% (95% CI, 5.1%–18.9%) for stroke, and 3.1% for trauma (95% CI, 0.9%–5.2%) cases. There was no significant change in 30-day mortality for cases, and no significant difference in change in mortality rates between cases and controls following the intervention. The difference in hospital bypass rates between cases and controls was 50.1% (95% CI, 42.3%–57.9%) for AMI, 36.2% (95% CI, 27.4%–44.9%) for stroke, and 32.2% (95% CI, 27.7%–36.8%) for trauma cases preintervention. Following the intervention, the difference in bypass rates dropped by 15.5% (95% CI, 3.5%–27.5%) for AMI, 25.3% (95% CI, 11.7%–38.8%) for stroke, and 22.7% (95% CI, 15.7%–29.6%) for trauma cases.
Conclusions:
Hospital closures did not affect patient mortality.
