Effect of the 9-11 terrorist attacks on suicide and homicide in the vicinity of the attacks

Objective: To describe the impact of September 11, 2001 on violent deaths (suicides and homicides) in the vicinity of the attacks. Methods: The daily mortality rate per 100,000 citizens was calculated for each attack site. Where the post-9-11 rate changed significantly, rates in concentric 150- and 300-mile geographic circles surrounding the site were also examined. To rule out the possibility that violent deaths were not investigated with the same care in the days after 9-11, selected control conditions were analyzed. To rule out the possibility that violent death rates may have changed for non-attack related reasons, corresponding 1998-2000 year rates were examined. For each mortality cause, if autocorrelation was present, a time series model was developed, incorporating a step function at 9-11 (intervention analysis--Box and Tiao, 1975). Where no autocorrelation was present, pre- post 9-11 rate difference was tested using a negative binomial distribution. Results: Within the New York City (NYC) Primary Metropolitan Statistical Area (PMSA), a statistically significant drop in suicides (but not homicides) was observed after 9/11/2001 (t=-2.8, p=.0046), while 1998-2000 analyses did not show corresponding effects. Counties within 150 miles of the NYC PMSA also showed a significant post-9-11 decrease in suicides. No corresponding difference in rates was found for Washington, DC or within 150 miles of the Flight 93 crash site. Conclusions: Both temporal and geographic aspects of suicide rates impacted by sentinel events such as 9-11 appear to be measurable, and such analyses may contribute valuable knowledge about how environmental forces mediate these rates.