The terrorist attacks in the United States on September 11, 2001 caused significant medical and psychiatric morbidity, particularly in Manhattan, and acutely disrupted the lives of people around the world. However, little research has examined the effects of September 11 outside the United States. This study hypothesized that the attacks influenced rates of deliberate self-harm, a complex behavioral phenomenon that includes deliberate self-poisoning. The authors conducted an ecological analysis of poisonings in the days immediately following September 11 in a population geographically removed from the events.
The authors identified all hospitalizations for self-poisoning in Ontario during the month of September from 1988 to 2003 using the population-based records of the Canadian Institutes of Health Information (International Classification of Diseases, Ninth Revision codes 960.0-990.0 and International Statistical Classification of Diseases, 10th Revision codes T36-T50.) To exclude instances of accidental toxicity, they restricted the analysis to poisonings in which the external cause of injury code indicated deliberate self-harm. They excluded more violent means of suicide (such as gunshot and hanging) because these are frequently fatal and may not lead to hospital admission. To examine the possibility that the results could reflect a generalized change in hospitalization or coding practices, they repeated the analyses using admissions for pneumonia, unstable angina, asthma, and congestive heart failure, diseases not governed by impulse.
The primary analysis examined admissions during the three-day period beginning on the second Tuesday in September of each year, corresponding to September 11-13, 2001. A three-day interval was studied because the authors anticipated an immediate but transient effect of the terrorist attacks and because research from Israel suggests that terrorist attacks reduce motor vehicle accidents for a similar period. In sensitivity analyses, the observation period following September 11 was changed from 3 days to 2 days and then to 4 days.
From 1988 to 2003, 6077 hospital admissions for deliberate self-poisoning in Ontario during September were identified. Of these, 614 occurred during the three-day period beginning on the second Tuesday of the month. In 2001, there were 13 admissions during this period, about 64% fewer than the predicted number of 36 (95% confidence interval, 25-46; P<.001). No similar phenomenon was seen when the analysis was repeated using the three-day period before (Saturday, Sunday, and Monday) or after (Friday, Saturday, and Sunday) the same three days in September. The findings persisted when the observation period was changed to 2 days or 4 days following September 11 (63% and 53% fewer than predicted, respectively; P = .002 for both comparisons). In contrast, the number of poisoning admissions in 2002 and 2003 did not differ significantly from predicted estimates. Men constituted a lower proportion of poisonings from September 11 to 13, 2001, compared with the same period in all other years (8% vs. 46%; P = .03). Repeating the analyses using admissions for pneumonia, unstable angina, asthma, and congestive heart failure showed no significant change in admissions for these conditions in the three days following September 11, 2001.
This study found that the terrorist attacks were associated with a transient but dramatic reduction in deliberate self-poisoning in Ontario, suggesting that some determinants of self-harm behavior can be temporarily suppressed by major world events.
