Background: Failure-to-rescue (FTR), originally developed to study quality of care in surgery, measures an institution's ability to prevent death after a patient becomes complicated.
Objectives: Develop an FTR metric modified to analyze acute myocardial infarction (AMI) outcomes.
Research design: Split-sample design: a random 20% of hospitals to develop FTR definitions, a second 20% to validate test characteristics, and an out-of-sample 60% to validate results.
Subjects: Older Medicare beneficiaries admitted to short-term acute-care hospitals for AMI between 2009 and 2011.
Measures: Thirty-day mortality and FTR rates, and in-hospital complication rates.
Results: The 60% out-of-sample validation included 234,277 patients across 1142 hospitals that admitted at least 50 patients over 2.5 years. In total, 72.1% of patients were defined as Medically Complicated (complex on admission or subsequently developed a complication or died without a recorded complication) of whom 19.3% died. Spearman r between hospital risk-adjusted 30-day mortality and FTR was 0.89 (P<0.0001); Mortality versus Complication=-0.01 (P=0.6198); FTR versus Complication=-0.10 (P=0.0011). Major teaching hospitals displayed 19% lower odds of FTR versus non-teaching hospitals (odds ratio=0.81, P<0.0001), while hospitals as a group defined by teaching hospital status, comprehensive cardiac technology, and having good nursing mix and staffing, displayed a 33% lower odds of FTR (odds ratio=0.67, P<0.0001) versus hospitals without any of these characteristics.
Conclusions: A modified FTR metric can be created that has many of the advantageous properties of surgical FTR and can aid in studying the quality of care of AMI admissions.