Disasters have been studied using a range of methodologies:
• Comparison of a representative sample of the exposed population with a control group [38,39].
• Examination of a subgroup of disaster victims, such as those seeking treatment or who are injured. Such studies often provide an opportunity for more in-depth analysis of some of the risk and protective factors .
• Investigation of the impact of disaster on particular groups, such as children  and emergency service personnel [42,43].
• Longitudinal cohort studies. Typically these studies have followed populations in the aftermath of an event, but there are several studies where a population has been studied for some other reason and then a disaster has impacted on that community [44-48].
• Identifying individuals in a general community sample who have had some disaster exposure and comparing them with people who have been exposed to other traumatic events or have had no traumatic exposure [36,37].
These different types of studies will often come to different conclusions about the same event, because of the issue of the representativeness of the sample . Also the time period that has elapsed between the event and the investigation can have a significant impact on the findings, because of the significant rates of natural remission in the first year.
The majority of the studies that have been published examined the impact of natural disasters (88 = 55%), a further 54 (34%) referred to technological disasters and 18 (11%) documented the impact of massive violence . These events occurred in 29 separate countries, with 57% having occurred in the United States. A further 29% of the events studied had occurred in Europe, Japan and Australia. The developing world -including Eastern Europe, Asia and Africa - accounted for only 14% of the studies.
A range of categories of victim populations has been studied. The vast majority of the survivors investigated were adults, while about 17% were school-aged children and adolescents [50,51]. A few studies have focused on emergency service personnel  and family assistance counsellors.
The majority of the studies provide a snapshot of the affected population within a 6 months window. Whilst there have been some substantial longitudinal studies which have followed populations for as long as 32 years , approximately half of the reports had their last data point less than one year after the event.
The methods used to measure morbidity varied considerably. Only a minority of studies used structured interviews [54-56], whereas the others used the more easily administered questionnaires [57,58]. A range of instruments has been used to characterise the psychological outcomes, with the most frequent focus being on PTSD . Patterns of non-specific morbidity have been often examined in populations, using instruments such as the General Health Questionnaire (GHQ) [60-63]. Other phenomena, such as disassociation and demoralisation, have also been studied [50,64]. The most comprehensive studies have used structured diagnostic interviews, such as the Composite International Diagnostic Interview (CIDI), allowing the generation of a series of psychiatric diagnoses . Approximately a quarter of the studies have also examined physical health concerns and problems [65-69]. Of particular interest in these studies has been the clinical worsening of physical symptoms and perceived illness burden in disaster-affected communities.
Was this article helpful?