Counting crises: US hospital evacuations, 1971-1999

To investigate the relative distribution of hazards causing hospital evacuations, thereby to provide rudimentary risk information for hospital disaster planning.

Cases of hospital evacuations were retrieved from newspaper and publication databases and classified according to hazard type, proximate and original cause, duration, and casualties. Both partial and full evacuations were included. The total number of evacuation incidents for all hazards were compared to the total number of hospital incidents for the one hazard, fire, for which national data is available.

There were 275 reported evacuation incidents from 1971-1999, with an annual average of 21 in the 1990s, the period for which databases were more reliable. The most, 33, were recorded in 1994, the year of the Northridge Earthquake. Of all incidents, 63 (23%) were attributable primarily to internal fire, followed by internal hazardous materials (HazMat) events (18%), hurricane (14%), human threat (13%), earthquake (9%), external fire (6%), flood (6%), utility failure (5%), and external HazMat (4%).

More than 50% of the hospital evacuations occurred because of hazards originating in the hospital facility itself or from human intruders. While natural disasters were not the preponderant causes of evacuations, they caused severe problems when multiple hospitals in the same urban area were incapacitated simultaneously. Clearly, as hospitals are vulnerable to many hazards, mitigation investments should be assessed not in terms of single-hazard risk-cost-benefit analysis, but in terms of capacity to mitigate multiple hazards. In view of the many qualifications and limitations of the dataset used here, but value of such data for disaster planning, hospitals should be asked to submit standardized incident reports to permit national data gathering on major disruptions.