Abstract 240: Regional Variation in 30-Day Ischemic Stroke Mortality and Readmissions in Get With the Guidelines-Stroke Hospitals

Abstract

Objective: To explore regional variation and correlation between 30-day ischemic stroke mortality and readmissions in ischemic stroke patients treated at Get With The Guidelines (GWTG)-Stroke hospitals. Methods: Hospital-level 30-day ischemic stroke mortality and readmission rates were generated by linking GWTG-Stroke national registry data (2007-11) to Medicare fee-for-service data. We then aggregated GWTG-Stroke hospitals into hospital referral regions (HRR) as defined by the Dartmouth Atlas. Frequency-weighted 30-day event rates were estimated for each HRR, which were risk-adjusted using HRR-level patient and hospital characteristics. A Pearson correlation coefficient (r) was used to estimate the linear relationship between mortality and readmission rates at the HRR level. Results: There were 429,631 ischemic stroke patients at least 65 years of age treated at 1,344 GWTG-Stroke hospitals in 282 different HRRs. The average HRR-level risk adjusted 30-day mortality rate was 10.3% (SD=1.1%), with range 7.4% to 15.6%. The average HRR-level risk adjusted readmission rate was 13.2% (SD=1.4%), with range 9.7% to 23.3%. HRR-level maps suggest that some regions with higher than average mortality - such as the Mountain West and Upper Midwest regions - had lower than average readmission rates. (Figure 1) There was a modest, but statistically significant negative correlation between 30-day mortality and readmission rates (r = -0.23, p<0.0001). (Figure 2) Conclusions: HRR-level 30-day ischemic stroke mortality and readmission rates exhibit substantial regional variation and are inversely correlated. Further research is ongoing to determine the origin of the regional variation and correlation between 30-day event rates.