Abstract 1: The Value of Adding Laboratory Data to Coronary Artery Bypass Graft Surgery Registry Data to Improve Models for Risk-Adjusting Provider Mortality Rates
Background: Clinical registry data for coronary artery bypass graft (CABG) surgery are currently being used in a few states and by the Society for Thoracic Surgeons to calculate provider risk-adjusted mortality rates. However, very little laboratory data are contained in these databases for purposes of risk-adjustment. The objective of this study was to examine whether adding laboratory data to New York’s CABG registry data will identify laboratory variables that are significant independent predictors of in-hospital/30-day mortality, and if so, whether the risk adjustment model containing these laboratory variables will significantly alter the quality assessments of New York hospitals where CABG surgeries are performed.
Methods: New York hospitals that performed CABG surgery in 2008-2010 were recruited into a study that required submitting data for 32 laboratory tests which were combined with CABG registry data. Statistical models were developed, with and without laboratory data, to compare risk-adjusted mortality rates. Discrimination, calibration, correlation in hospital risk-adjusted mortality rates, and differences in hospital quality outlier status were examined for both registry and registry/laboratory models.
Results: There were 9,958 patients who underwent isolated CABG surgery at 15 participating hospitals during the study period. Discrimination of statistical models was very similar (C = 0.783 for the registry model and 0.812 for the registry/laboratory model, P = 0.26). Most of the non-laboratory variables in the two models were identical, except that the registry model contained diabetes requiring medication and the registry/laboratory model included recent congestive heart failure and renal dialysis. The registry/laboratory model also included two laboratory variables: Aspartate Aminotransferase (AST) > 50 U/L, and International Normalized Ratio (INR) >1.5. The addition of laboratory data did not affect the quality outlier status of hospitals, and had minimal effect on the tertile of risk-adjusted mortality of New York hospitals.
Conclusion: Clinical models with and without laboratory data had similar discrimination, hospital risk-adjusted mortality rates were essentially unchanged, and hospital outlier status was identical. Two laboratory variables, AST and INR, were significant independent predictors of mortality and might be valuable additions to CABG clinical registry data. The benefits and practical challenges of adding key laboratory variables to CABG registry data should be carefully considered.
Author Disclosures: F. Qian: None. E.L. Hannan: None. M. Pine: None. B.A. Dennison: None.
- © 2014 by American Heart Association, Inc.