Send to

Choose Destination
Med Care. 2017 May;55(5):476-482. doi: 10.1097/MLR.0000000000000691.

Differences in Hospital Risk-standardized Mortality Rates for Acute Myocardial Infarction When Assessed Using Transferred and Nontransferred Patients.

Author information

*Division of Pulmonary, Allergy, and Critical Care Medicine †Department of Critical Care Medicine, CRISMA Center, University of Pittsburgh School of Medicine ‡Department of Health Policy & Management, University of Pittsburgh Graduate School of Public Health §Clinical and Translational Science Institute, University of Pittsburgh ∥Division of Cardiology, University of Pittsburgh School of Medicine, Pittsburgh, PA.



One in 5 patients with acute myocardial infarction (AMI) are transferred between hospitals. However, current hospital performance measures based on AMI mortality exclude these patients from the evaluation of referral hospitals.


To determine the relationship between risk-standardized mortality for transferred and nontransferred patients at referral hospitals.


This is a retrospective cohort study.


Fee-for-service Medicare claims from 2011 for patients hospitalized with a primary diagnosis of AMI, at hospitals admitting at least 15 patients in transfer.


Hospital-specific risk-standardized 30-day mortality rates (RSMRs) for 2 groups of patients: those admitted through transfer from another hospital, and those natively admitted without a preceding or subsequent interhospital transfer.


There were 304 hospitals admitting at least 15 patients in transfer. These hospitals cared for 77,711 natively admitted patients (median, 254; interquartile range, 162-321), and 11,829 patients admitted in transfer (median, 26; interquartile range, 19-46). Risk-standardized mortality rates were higher for natively admitted patients than for those admitted in transfer (mean, 11.5%±1.2% vs. 7.2%±1.1%). There was weak correlation between hospital performance as assessed by RSMR for patients natively admitted versus those admitted in transfer (Pearson r=0.24, P<0.001); when performance was arrayed by quartile, 102 hospitals (33.6%) differed at least 2 quartiles of performance across the 2 patient groups.


For Medicare patients with AMI, hospital-specific RSMRs for natively admitted patients are only weakly associated with RSMRs for patients transferred in from another hospital. Current AMI performance metrics may fail to provide guidance about hospital quality for transferred patients.

[Indexed for MEDLINE]
Free PMC Article

Supplemental Content

Full text links

Icon for Wolters Kluwer Icon for PubMed Central
Loading ...
Support Center