Abstract 14288: Substantial Hospital-level Variation in Troponin Testing After Non-cardiac Surgery

Abstract

Introduction: There is an increasing emphasis in detecting postoperative myocardial infarction (PMI) using routine troponin testing after non-cardiac surgery. However, clinical practice guidelines vary considerably in their recommendations. We conducted a population-based cohort study in Ontario, Canada to assess the degree of hospital-level variation associated with troponin testing after three commonly performed surgeries. Methods: We conducted a retrospective cohort study of adults (40-105 y) undergoing major orthopedic, colorectal, or vascular surgery in Ontario, Canada from January 1, 2010 to December 31, 2017. Hierarchical logistic regression modeling was used to model the association of patient, surgery, and hospital factors with postoperative troponin testing, while accounting for clustering at the hospital level. Results: We identified 176,454 eligible patients. Canadian Cardiovascular Society guidelines recommended troponin testing for 73.5%, 90.8% and 95.6% of orthopedic, colorectal, and vascular surgery patients respectively, but only 6.7%, 16.6%, and 50.2% were actually tested. Inter hospital variation in testing rates was considerable for the three surgeries (Figure; 0-33%, 0-38% and 18-84%). Even after risk-adjustment, the median odds ratio for testing across hospitals was still 1.74, 1.63, and 2.65 for orthopedic, colorectal, and vascular surgery, respectively. This corresponded to intraclass correlation coefficients of 9.3%, 7.4%, and 24.2% respectively. Conclusion: Despite strong recommendations by Canadian guidelines for troponin testing after non-cardiac surgery, testing rates were low overall and varied significantly across hospitals.