Oversight May Reduce Military Health System Leakage: Preventing Inappropriate Interhospital Transfers

Abstract

Introduction: The Military Health System (MHS) offers an example of a socialized healthcare model, operating within a larger "purchased care"civilian healthcare market. This arrangement has facilitated a trend wherein MHS clinicians often transfer moderate-to-complex patients to surrounding civilian hospitals, despite having the capability to care for such patients in-house. In an effort to stem this behavior, two initiatives were introduced at Carl R Darnall Army Medical Center (CRDAMC): A Transfer Policy Statement and Transfer Rounds. The Transfer Policy Statement emphasized that patients ought to be transferred only for capability gaps within the hospital. Transfer Rounds were then used to review the care received by each transferred patient and assess if that care could have been delivered internally. The purpose of this study is to assess the effect of these initiatives on reducing transfers from our hospital. Materials and Methods: We performed a retrospective chart review from July 2019 through June 2020 to identify the number of total emergency department (ED) transfers, subcategorized as either transfers we had the capability to care for or transfers we did not have the capability to care for. The Transfer Policy Statement was published in August 2019, and Transfer Rounds were instituted in November 2019. We hypothesized that the two interventions would decrease the number of monthly inappropriate transfers. This was assessed by analyzing the proportion of inappropriate to appropriate patient transfers via Cochran and Armitage using SAS 9.4 (SAS Institute, Cary, NC). The projected received an Exemption Determination from the CRDAMC's Human Research Protections Office. The Defense Health Agency approved the data-sharing agreement. Results: Over the study period, a total of 706 transfers met the criteria for analysis. The monthly median for total ED transfers was 64.5 (Interquartile Range (IQR) 45-74); appropriate transfers averaged 29.5 (median, IQR 24.5-36) and inappropriate 25.5 (median, IQR 9-41.5). A statistically significant downward trend in the fraction of inappropriate transfers was demonstrated by Cochran and Armitage (P