Purpose Regionalization of complex surgery to high-volume hospitals has been advocated based on cross-sectional volume-outcome studies. In April 2007, the agency overseeing cancer care in Ontario, Canada, implemented a policy to regionalize lung cancer surgery at 14 designated hospitals, enforced by economic incentives and penalties. We studied the effects of implementation of this policy. Methods Using administrative health data, we used interrupted time series models to analyze the immediate and delayed effects of implementation of the policy on the distribution of lung cancer surgery among hospitals, surgical outcomes, and health services use. Results From 2004 to 2012, 16,641 patients underwent surgery for lung cancer. The proportion of operations performed in designated hospitals increased from 71% to 89% after the policy was implemented. Although operative mortality decreased from 4.1% to 2.9% (adjusted odds ratio, 0.68; 95% CI, 0.58 to 0.81; P < .001), the reduction was due to a preexisting declining trend in mortality. In contrast, in the years after implementation of the policy, length of hospital stay decreased more than expected from the baseline trend by 7% per year (95% CI, 5% to 9%; P < .001), and the distance traveled by all patients to the hospital for surgery increased by 4% per year (95% CI, 0% to 8%; P = .03), neither of which were explained by preexisting trends. Analyses limited to patients ≥ 70 years of age demonstrated a reduction in operative mortality (odds ratio, 0.80 per year after regionalization; 95% CI, 0.67 to 0.95; P = .01). Conclusion A policy to regionalize lung cancer surgery in Ontario led to increased centralization of surgery services but was not independently associated with improvements in operative mortality. Improvements in length of stay and in operative mortality among elderly patients suggest areas where regionalization may be beneficial.
Variation in length of hospital stay after lung cancer surgery in the Netherlands†
