Use of 90-day mortality does not change assessment of hospital quality after coronary artery bypass grafting in New York State

Abstract

Objectives: Publicly reported postoperative 30-day mortality rates are commonly used to compare hospital quality after coronary artery bypass graft (CABG) surgery. We sought to determine whether 90-day mortality rates, which are not publicly reported but better capture postdischarge mortality, are a better determinant of hospital performance. Methods:Retrospective cohort analysis of 30 versus 90-day risk-standardized mortality rates (RSMR) at adult cardiac surgical centers in NYS from 2008 -2014. Hospitals were classified as good or poor-performing outliers at each time point based on the bounds of the 95% confidence interval around each hospital's predicted RSMR, determined via hierarchical models. The primary outcome was change in institutional performance via outlier classification from 30 to 90 days.Results: During the study period, 72,398 adults underwent a CABG procedure at one of 42 institutions. The RSMR increased from 30 to 90 days at all institutions, with a median 30-day RSMR of 2.16% (IQR 0.69%) and median 90-day RSMR of 3.69% (IQR 1.00%). In using a 90 instead of 30-day metric, 3 hospitals changed outlier status. One hospital improved to a good from as expected performer; 2 worsened to as expected from good performers. Conclusions:In a cohort of patients who underwent CABG surgery from 2008 -2014 in NYS, use of a 90-day mortality metric resulted in a change in hospital quality assessment for a minority of hospitals. The use of 90-day mortality may not provide additional value when evaluating institutional performance for this population.