Amesh A. Adalja, MD, FACP, FACEP, March 20, 2015
Where do emergency department (ED) patients go for care when a hospital is closed as a result of a disaster? Does the pattern of use differ between public and private hospitals? These are questions explored in an important new study of how ED use patterns changed for users of 2 closed EDs after Hurricane Sandy. The study by Lee et al is published in Disaster Medicine and Public Health Preparedness.
Hurricane Sandy, which struck many parts of the eastern United States in 2012, had a particularly strong impact on the hospitals of New York City as it forced several hospital evacuations, including 2 major facilities in Manhattan: the public (city government–owned) Bellevue Hospital Center and the private New York University Langone Medical Center (NYU). The EDs of these 2 hospitals remained closed for several months, creating a surge of ED patients at other NYC hospitals.
Redistributed Patients Identified
Using a database of ED visits, Lee et al identified cohorts of patients who were established exclusive users of one or the other of the closed EDs (Bellevue or NYU) and tracked their post-Sandy ED visits. Before the hurricane, Bellevue’s ED patients tended to be younger, nonwhite, and less likely to have private insurance than NYU’s patients. Additionally, many established Bellevue ED patients traveled from the outer boroughs of NYC to the hospital, whereas NYU’s ED patients tended to cluster in Manhattan.
Several important findings included:
The events studied in this paper highlight the fact that while many patients used the nearby Beth Israel Medical Center after Hurricane Sandy, many—and especially those who relied on Bellevue’s ED—did not. Post-hurricane, many displaced Bellevue ED patients, who before the hurricane may have traveled past many hospitals to Bellevue’s ED, preferred to use other public hospitals rather than nearer private alternatives.
The results of this fascinating study should inform emergency preparedness planning. Many hospitals have unique patient populations that are not always interchangeable. This might be because the hospital offers unique services (eg, psychiatry or pediatrics) or is known to accommodate certain populations (eg, non-English speakers) particularly well. Particular payment sources (eg, veterans’ benefits or payments for medically indigent patients) could also influence patients’ hospital preferences.
These results should also inform healthcare coalition building. This study suggests that, as hospital emergency preparedness coalitions are evolving, they should take into consideration patient variables other than simple geographic proximity and medical service availability (eg, trauma care). For example, the medical centers of the Department of Veterans Affairs often participate in geographic hospital coalitions as well as in coalitions of other veterans’ hospitals (which may be geographically distant). Such a practice acknowledges that patient populations may prefer similar hospitals even if they are not the closest, and it is important for hospitals to be ready in the event of a disaster.
Lee DC, Smith SW, Carr BG, Goldfrank LR, Polsky D. Redistribution of emergency department patients after disaster-related closures of a public versus private hospital in New York City. Disaster Med Public Health Prep 2015;17:1-9.