Quality Measures and Institutional Factors Are Not Associated with Increased Inter‐Hospital Transfer Mortality

1University of Vermont College of Medicine, Burlington, VT

Meeting: Hospital Medicine 2014, March 24-27, Las Vegas, Nev.

Abstract number: 100

Background:

Hospital mortality rates are frequently used as indicators of quality. Patients undergoing inter‐hospital transfer represent a unique subset of patients that contribute to hospitals’ overall mortality rates. Factors associated with mortality rates following inter‐hospital transfer have not been clearly delineated.

Methods:

We conducted a cross sectional study of 104 academic medical centers (AMC) that comprised the principal members in the University Health System Consortium (UHC) database. Observed institutional mortality rates for patients transferred from another hospital in 2012 were compared with publicly reported quality measures and institutional factors. Heart failure, pneumonia, and acute myocardial infarction process measure data for AMC with reporting rates sufficient for analysis were obtained from the Hospital Compare database for the time period from October 1, 2011 to September 30, 2012. Institutional characteristics were obtained from the UHC database. Regional density of acute care hospitals was obtained from the Hospital Compare database and Google Maps.

Results:

The mean observed mortality rate among transfer patients (5.60%; 95% CI 5.11, 6.08) was significantly higher than the mean observed mortality rate among non‐transfer patients (1.64%; 95% CI 1.55, 1.73) with a p value of <0.001. Transfer mortality did not correlate with non‐transfer mortality or with Case Mix Index (Table 1). There was no positive correlation between institutional performance on diagnosis‐specific process measures and transfer mortality (Table 1). Institutional and regional characteristics were not associated with transfer mortality (Table 1).

Conclusions:

Patients undergoing inter‐hospital transfer have a dramatically higher mortality rate than non‐transfer patients. The lack of association between transfer mortality and non‐transfer mortality suggests that the factors impacting patient outcomes in these two populations may differ. Furthermore, measures of diagnosis‐specific processes of care were not positively associated with transfer mortality. Without robust risk adjustment, the inclusion of transfer mortality in hospital quality measurement may be misleading.

The study has notable limitations. First, data sets were obtained for overlapping but non‐identical time periods. Second, the results may be skewed by outlier institutions with few transfer patients and transfer mortality rates highly vulnerable to individual deaths. As example, the removal of one such outlier institution from the data set altered a number of the correlations reported. Detailed elaboration of these outlier effects is needed. Further work is clearly warranted to understand the factors underlying mortality in transferred patients.

To cite this abstract:

Borofsky J, Roberts J, Repp A. Quality Measures and Institutional Factors Are Not Associated with Increased Inter‐Hospital Transfer Mortality. Abstract published at Hospital Medicine 2014, March 24-27, Las Vegas, Nev. Abstract 100. Journal of Hospital Medicine. 2014; 9 (suppl 2). https://www.shmabstracts.com/abstract/quality-measures-and-institutional-factors-are-not-associated-with-increased-interhospital-transfer-mortality/. Accessed November 17, 2019.

« Back to Hospital Medicine 2014, March 24-27, Las Vegas, Nev.