Article ID: | iaor201525286 |
Volume: | 4 |
Issue: | 1 |
Start Page Number: | 29 |
End Page Number: | 40 |
Publication Date: | Mar 2015 |
Journal: | Health Systems |
Authors: | Zhang Kun, Howard David H |
Keywords: | simulation, statistics: regression |
Simulation and surge studies benefit from realistic assumptions about changes in patient flows during disasters. We sought to determine the impact of Hurricane Katrina and the 2008 Midwest floods on hospital and skilled nursing facility (SNF) discharge patterns. We used Medicare claims to identify beneficiaries admitted to hospitals or SNFs in the affected areas for each disaster. We used Cox proportional hazards regression with time‐varying covariates for disasters to assess the impact of disasters on the likelihood that a patient would be discharged. Discharges from SNFs and hospitals declined in the week following Hurricane Katrina. Discharges from SNFs declined in the 3‐week period during the Midwest floods but hospital discharge rates were unaffected. Reductions in discharge rates may reduce the ability of health‐care facilities to create surge capacity during disasters.