Temporal Variations in School-Associated Student Homicide and Suicide Events — United States, 1992-1999

Morbidity and Mortality Weekly Report, August 10, 2001

Recent, widely reported violent deaths associated with schools have led many adults to believe that a school shooting could occur in their community and many children to express increasing concern about their own safety at school [1]. CDC, in collaboration with the U.S. Education and Justice departments, has been tracking school-associated violent deaths [*] since the 1992-1993 school year [2]. To evaluate whether the risk for school-associated violent death varies during the school year, CDC analyzed monthly counts of school-associated homicide and suicide events that occurred among students in elementary and secondary (middle, junior high, and senior high) schools in the United States. This report summarizes the results of these analyses, which indicate that student homicide event rates are usually highest near the start of the fall and spring semesters, and suicide event rates are highest during the spring semester. These findings can assist school personnel in planning and implementing violence-preventio n programs.

For these analyses, a school-associated violent death event was defined as a homicide or suicide of a student in which the fatal injury occurred 1) on the campus of a functioning public or private elementary or secondary school in the United States, 2) while the victim was on the way to or from regular sessions at such a school, or 3) while the victim was attending or traveling to or from an official school-sponsored event. Events resulted in the death of at least one student but may have included the deaths of nonstudents (e.g., faculty, school staff, family members, and community residents). Events were identified through a systematic search of two computerized newspaper and broadcast media databases (Lexis-Nexis and Dialog) [3,4]. To confirm events, a qualifying interview was conducted with at least one law enforcement or school official familiar with each event.

Student homicide and suicide event rates were analyzed individually for the 10 months that define a typical school year (September-June). Events that involved the homicide of a student followed by the suicide of a student perpetrator were included in each analysis. Event totals for each month were calculated by summing over the 7 school years in the study period. For both homicide and suicide events, the relevant exposure period in each month was based on the total number of school days in that month over the entire 7-year period, estimated by inspection of several school calendars selected from each region of the country. For each event type, the number of events per school day was calculated for each month in the school calendar and plotted to allow visual assessment of trends.

Poisson rate models were used to evaluate the trends over the school year. Each model was restricted to one monthly time-trend variable and one semester transition variable to account for the apparent increase in event rates following the semester/holiday break that usually occurs in late December through early January.

For the 7 school years during September 1, 1992-June 30, 1999, 209 school-associated violent death events occurred that involved either the homicide or suicide of a student. During the 7 school years of the study period, an average of 0.14 school-associated homicide events occurred each school day (one event every 7 school days) (Figure 1). For homicide events, rates decreased during each semester (monthly change in log rate: -0.2; p=0.0002) and increased markedly in association with the transition between the fall and spring semesters (increase in log rate: 0.98; p=0.001). These findings indicate that homicide event rates were relatively high near the beginning of the school year, gradually declined during the fall semester, and exhibited a similar pattern during the spring semester.

For suicides, an average of 0.03 events occurred each school day (one event every 31 school days). The estimated Poisson rate model for suicide events involved a nonsignificant time-trend variable. As a result, this variable was subsequently dropped and the resulting simplified model, which included only the semester transition variable, suggests that the suicide event rate was higher during the spring semester than the fall semester (increase in log rate: 1.0; p=0.0103).

Reported by: Safe and Drug Free Schools Program, US Dept of Education. National Institute of Justice, US Dept of Justice. Div of Violence Prevention and Office of Statistics and Programming, National Center for Injury Prevention and Control, CDC.

Editorial Note: The findings in this report suggest significant systematic temporal variations in school-associated student homicide and suicide events. Student homicide event rates were highest near the start of each semester and then declined over the following months. In comparison, suicide event rates did not show any significant variation within semesters, but the overall rate was significantly higher in the spring semester than in the fall semester.