especially clinical factors, are surveyed, but often not through population-based surveys that would avoid the bias and lack of generalizability of clinical populations (Feinstein, 1977). Most information regarding suicide attempts must be collected from data systems designed for other purposes (PHS, 2001). This section describes a few of the potential models and sources of information on suicide attempts: the National Electronic Injury Surveillance System, the Youth Risk Behavior Survey, and the Oregon State Adolescent Suicide Attempt Data System.
The NEISS has been operated by the U.S. Consumer Product Safety Commission (CPSC) for almost 30 years. In 2000, the system was expanded to collect data on all injuries, and since 1992 NEISS has collected information on all nonfatal firearm-related injuries seen in NEISS emergency rooms (Annest et al., 1995; Davis et al., 1996); some of these incidences may represent suicide attempts. NEISS is based on injury data gathered from the emergency departments of 100 representative hospitals selected as a probability sample of all 5,300+ U.S. hospitals with emergency departments (EDs) (grouped into 5 “strata,” four representing EDs of differing sizes and one from children’s hospitals) (CPSC, 2001).
Given its current sampling system, the utility of NEISS is limited; the data can only be used for national estimates and are invalid at regional, state and local levels (GAO, 1997). In addition, because it does not use the International Classification of Diseases (ICD) coding system both the detail of data collected and the ease with which the data can be shared with other systems are limited (AdvanceMed, 2001).
The YRBS is managed by the CDC and includes national, state, territorial, and local school-based surveys of representative samples of students in grades 9–12 in participating jurisdictions. Its intent is to monitor risk behaviors associated with the leading causes of injury and death among adolescents (Kann et al., 1998). In 1997, the YRBS was conducted by 38 states, 4 territories and 17 large cities, in addition to the national-level representative survey (STIPDA, 1999). Concerns regarding the validity of self-reports presents particular problems for collecting information on suicide attempts by the YRBS (Ivarsson et al., 2002). In addition, some jurisdictions, especially less populous ones, choose not to include the items about suicidality out of concerns for liability and imitation. This introduces bias into the results when comparing geographical areas. How-