Association Between Hospitals' Risk-Adjusted Emergency Department Visits and Survival and Costs in Kidney Cancer Patients Undergoing Nephrectomy.

Abstract

To estimate the association between a hospital's risk-adjusted emergency department (ED) visit rate and its risk-adjusted mortality rate and costs among kidney cancer patients undergoing initial nephrectomy. Using 2007-2012 Surveillance, Epidemiology, and End Results (SEER)-Medicare data, we used logistic regression to model ED visit occurrence within 30 and 365 days for all kidney cancer patients receiving initial surgery. Our model controlled for demographics, stage, histology, systemic targeted therapy, and comorbidities. Based on model predictions, we created a ratio of actual versus predicted ED visits for hospitals to identify hospitals with higher and lower than predicted ED visit rates. We estimated the association between the hospitals' ED visit ratio and hospitals' risk-adjusted 365-day mortality rates, and 6- and 12-month total costs and total costs (less ED visits). In our sample of 6078 patients, 15.5% had an ED visit within 30 days of surgery and 43.5% within 365 days. For hospitals with≥ 11 patients, we found no statistically significant association between 30-day or 365-day risk-adjusted ED visit rate and their 365-day risk-adjusted mortality rate. Hospitals' 30-day ED visit rates were not significantly associated with either 6- or 12-month costs. However, hospitals' 365-day ED visit rates were significantly associated with 12-month costs, even when excluding the cost of the ED visit. Our results suggest hospitals' risk-adjusted ED visit rates capture a qualitatively different measure of quality than the more commonly reported mortality rates. Longer term ED visit rates are significantly associated with increased costs while 30-day ED visits are not.