The earliest disaster studies were case studies of specific events. Parallel with developments in research methods in epidemiology generally, particularly case-control methodologies, disaster researchers began to apply the basic principles of epidemiology to understanding the psychological aftermath of these events. For example, the case-control design was recently applied by Asukai  in a study of PTSD among firefighters who had been called in as rescue workers after the Sarin nerve gas attack in the Tokyo subway. They found that PTSD occurred more frequently in firefighters hospitalized with signs of Sarin poisoning, while most non-PTSD subjects had no, or only mild, intoxication.
Disasters can also be conceptualized as "natural experiments". Like experimental research, disaster studies often involve careful comparisons of exposed and non-exposed cases. Indeed the design of the prototypical disaster study can most readily be understood as an extension of a modified cohort design. Most of the time, however, we do not know baseline morbid ity level of the population affected by the disaster and must infer the level from epidemiologic studies of demographically similar populations.
There have been some notable exceptions, however, in which a population happened to have been studied prior to the disaster, and then a postdisaster follow-up was performed. One example involved a population in Puerto Rico who had participated in a psychiatric epidemiologic study modeled on the Epidemiologic Catchment Area (ECA) study  in 1984 . The following year, torrential rains hit the island, causing extensive mudslides and leaving 180 people dead, 4000 in shelters, and 19000 with serious property damage. In 1987, the investigators re-evaluated a group of disaster survivors (n = 77) and controls (n = 298), using a Spanish-language version of the Diagnostic Interview Schedule Disaster Supplement . At approximately the same time, an area in St. Louis, Missouri, that was also part of the ECA study, was struck by a series of disasters. First devastating floods swept through the area, causing five deaths and necessitating the evacuation of about 25000 people. Evacuees from the residential area of Times Beach were later informed that they could not return to their homes because of the contamination of soil and water with dioxin. The discovery of a radioactively contaminated water well and more floods and a tornado followed over the next few months. In this study, pre- and post-disaster interviews were conducted on 44 subjects exposed either to flood alone, dioxin, tornadoes, radioactively contaminated water, or multiple events (n's for the different exposures ranged from 2 to 20). The results showed that there were significant increases in depressive, somatic, and PTSD symptoms in the affected populations in both Puerto Rico [33-35] and St. Louis [36,37]. In another example, Trevisan et al.  were conducting a 5-year follow-up of Italian workers participating in a study of coronary heart disease risk factors when an earthquake struck. Workers whose examinations were performed afterward had significantly higher heart rates, serum cholesterol levels, and triglyceride levels than those whose evaluations were completed before the earthquake. Yet another example involved a study of depressive and post-traumatic stress symptoms after an earthquake in California. In this study, a group of university students had completed a set of questionnaires 10 days before the earthquake struck. They were re-evaluated 7 weeks post-disaster. The authors found that pre-event symptomatology was the strongest predictor of post-disaster symptom severity .
Was this article helpful?