Abstract TP322: Additional Factors May Affect Hospital Performance Ratings on Ischemic Stroke Care in California

Abstract

Objectives: A report on ischemic stroke care at California hospitals was based on risk-adjusted 30-day mortality and accounted for many important risk factors (patient demography, clinical characteristics, and stroke severity). Other studies have indicated that geographic location, hospital characteristics and insurance type may also be drivers of differences in quality of care. The effect that these and other factors may have on hospital performance ratings for ischemic stroke is not well understood. Methods: Data used were from the California patient discharge data 2011-2013, collected by the Office of Statewide Health Planning and Development (OSHPD). Hospitals were chosen according to their performance in two recent hospital “report cards” on stroke produced by OSHPD. We compared 15 hospitals rated as “Better” with 14 “Worse” hospitals (10615 patients) on patient demographics, geographic location, insurance type, hospital characteristics, tissue-type plasminogen activator (tPA) use and expected 30-day mortality rate. Results: Patients admitted to “Worse” hospitals were more likely to be younger, white or Hispanic, and reside in lower income zip codes than “Better” hospitals ( P <0.001). “Worse” hospitals served a significantly higher percentage of patients with Medi-Cal insurance than “Better” hospitals (14.4% vs 9.6%, P <0.001). There were no significant differences in hospital geography or teaching status, bed size or Get With the Guidelines-Stroke Hospitals status between “Worse” and “Better” hospitals. Patients admitted to “Worse” hospitals had similar lengths of stay as those at “Better” hospitals and the transfer rate was also similar. “Worse” hospitals coded significantly fewer secondary diagnoses compared to “Better” hospitals (40.5% vs 53.0%, P <0.005). The tPA usage rate was significantly higher in the “worse” group than the “Better” group (11.5% vs 9.2%, P <0.005). “Worse” hospitals had significantly lower expected 30-day mortality rates compared to “Better” hospitals (8.8% vs 11.6%, P <0.005). Conclusion: Hospital performance ratings on ischemic stroke outcome were significantly associated with patient geographic location, socioeconomic status, and insurance type, but were not related to hospital characteristics.