Abstract TP48: Acute Ischemic Stroke Transfer Networks: Low Volume Ed-hospital Connections Are Common

Abstract

Background: Emergency Department (ED) transfers to tertiary centers are a critical component of stroke systems of care, yet little is known about the status of the transfer networks used to execute this process. Our objective is to characterize these transfer networks and assess rates of acute treatments among transfers to identify potential targets for improving organization of stroke systems of care. Methods: Using the 2017 US Centers of Medicare and Medicaid Inpatient and Outpatient Fee-for-Service files, we captured patients with inpatient ischemic stroke claims based on ICD 10 codes. We defined ED-hospital transfers as: 1) any outpatient ED claim ≤48 hrs prior to the inpatient stroke claim, 2) ED discharge status indicating an acute care hospital, and 3) originating and receiving facilities with different NPI numbers. We characterized the structure of transfer networks and acute treatment rates (based on ICD 10 codes and DRGs for thrombolysis and thrombectomy) using descriptive statistics. Results: We identified 225,177 patients with inpatient stroke claims. Of these, 7,907 (3.5%) had an ED-hospital transfer, involving 2083 different acute care hospitals. Amongst EDs that transferred a patient, the median number of receiving hospitals was 1 (IQR 1; range [1,7]) and the median volume of transferred patients was 3 (IQR 5; range [1,53]). Amongst hospitals that had stroke patients transfer in, the median number of originating EDs was 2 (IQR 5; range [1,28]) and the median volume of transferred patients was 5 [IQR 14; 1-158]. Amongst all ED-hospital transfer pairs, the median volume of transferred patients was 2 (IQR 2; range[1,53]). Amongst transferred patients, 60% received thrombolysis and 16% received mechanical thrombectomy. Conclusions: We found wide variation in the volumes of ED-hospital transfers for acute stroke, with a large number of these transfers occurring between hospitals with low transfer volumes. Given that high volume centers generally have better outcomes and the majority of identified transfers received acute therapies, a better understanding of the clinical circumstances driving these low-volume transfer connections is needed.