Patient characteristics associated with undergoing cancer operations at low-volume hospitals

Background. Although strong volume-outcome relationships exist for many cancer operations, patients continue to undergo these operations at low-volume hospitals. Methods. Patients were identified from the National Cancer Data Base from 2010-2013 who underwent resection for bladder, breast, esophagus, lung, pancreas, rectum, and stomach cancers. Low volume hospitals were defined as those in the bottom quartile by surgical volume for each cancer type separately. Logistic regression models were constructed to assess patient-level factors associated with undergoing cancer surgery at low-volume hospitals across cancer types while controlling for tumor characteristics. Survival outcomes (30- and 90-day mortality; overall survival) were also assessed. Results. Low volume thresholds were 4, 84, 4, 18, 8, 7, and 4 resections per year for bladder, breast, esophagus, lung, pancreas, rectum, and stomach cancers, respectively, resulting in 772 (74.1 %), 828 (57.5%), 664 (77.5%), 830 (64.7%), 716 (79.2%), 898 (65.1%), and 888 (68.5%) hospitals classified as low-volume hospitals, respectively. For all the cancers examined, patients were more likely to undergo operation at low-volume hospitals if they traveled shorter distances (home to surgical facility), resided in rural locations, or had not received neoadjuvant therapy. Other patient and tumor factors were not associated consistently with undergoing operation at low-volume hospitals. Patients who went to low-volume hospitals had poorer outcomes among the studied cancers. Conclusion. Patients continue to undergo operation at low-volume hospitals due to where they live and how far they have to travel. Regionalization policy initiatives will remain challenging in this population. Efforts should therefore continue to emphasize quality improvement locally at each facility caring for patients with cancer.