Potentially avoidable inter-facility transfer from Veterans Health Administration emergency departments: A cohort study

Abstract

Background: Inter-facility transfer is an important strategy for improving access to specialized health services, but transfers are complicated by over-triage, under-triage, travel burdens, and costs. The purpose of this study is to describe ED-based inter-facility transfer practices within the Veterans Health Administration (VHA) and to estimate the proportion of potentially avoidable transfers.

Methods: This observational cohort study included all patients treated in VHA EDs between 2012 and 2014 who were transferred to another VHA hospital. Potentially avoidable transfers were defined as patients who were either discharged from the receiving ED or admitted to the receiving hospital for ≤1 day without having an invasive procedure performed. We conducted facility- and diagnosis-level analyses to identify subgroups of patients for whom potentially avoidable transfers had increased prevalence.

Results: Of 6,173,189 ED visits during the 3-year study period, 18,852 (0.3%) were transferred from one VHA ED to another VHA facility. Rural residents were transferred three times as often as urban residents (0.6% vs. 0.2%, p < 0.001), and 22.8% of all VHA-to-VHA transfers were potentially avoidable transfers. The 3 disease categories most commonly associated with inter-facility transfer were mental health (34%), cardiac (12%), and digestive diagnoses (9%).

Conclusions: VHA inter-facility transfer is commonly performed for mental health and cardiac evaluation, particularly for patients in rural settings. The proportion that are potentially avoidable is small. Future work should focus on improving capabilities to provide specialty evaluation locally for these conditions, possibly using telehealth solutions.

Keywords: Emergency service, hospital; Hospitals, rural; Regionalization; Rural health services; Veterans health.

Fig. 1

Flow diagram of study participants

Fig. 2

Distribution of inter-facility transfers by Clinical Classification Software (CCS) diagnosis group. Each bar shows the number of transfers within each diagnosis group. The left bar (black and white), shows the number of transfers to Veterans Health Administration (VHA) facilities, stratified by potentially avoidable transfer (PAT) status (left vertical axis). The right bar shows the number of non-VHA transfers (right vertical axis). The relative height of the black/white bar and the grey bar shows compares the distributions in transfers to VHA facilities vs. non-VHA facilities. Categories (horizontal axis) are CCS categories, with the CCS category number listed in parentheses after each category

Fig. 3

Map of ED-based VHA-to-VHA inter-facility transfers, 2012–2014. Each dot on the map indicates a single VHA hospital with an emergency department (ED). Lines between these hospitals indicate transfers between facilities, with the thickness of the line represents the number of transfers. For some pairs of hospitals, the number of transfers are bidirectional, in which case the number of transfers in each direction are added together to represent the total transfer volume. Lines are not drawn between hospitals that have fewer than 100 VHA-to-VHA transfers over the study period. The proportion of transfers within each Veterans Integrated Service Network (VISN) that qualify as potentially avoidable transfers (PAT) is represented by grayscale shading (see legend). Note that non-VHA transfers are not included on this figure. The authors would like to acknowledge Morgan Swanson, BS for her assistance with preparation of this figure

Fig. 4

Regional variation in potentially avoidable transfers (PATs) by Clinical Classification Software (CCS) diagnosis group. Each cell in the heat map represents the total number of potentially avoidable transfers within one Veterans Integrated Service Network (VISN) region. CCS diagnosis categories are listed on the vertical axis, with the diagnosis group number listed in parentheses after the CCS category abbreviation. VISN regions are listed on the horizontal axis. Darker colors represent more PATs within the VISN for the diagnosis group