National Strategy for Suicide Prevention:
Goals and Objectives for Action

GOAL 11:
Improve and Expand Surveillance Systems

Surveillance has been defined as the systematic and ongoing collection of data (Bonnie et al., 1999). Surveillance systems are key to health planning. They are used to track trends in rates, to identify new problems, to provide evidence to support activities and initiatives, to identify risk and protective factors, to target high risk populations for interventions, and to assess the impact of prevention efforts (Thacker & Stroup, 1994).

Data are needed at national, state and local levels. National data can be used to draw attention to the magnitude of the suicide problem and to examine differences in rates among groups (e.g., ethnic, age groups) and locales (e.g., rural vs. urban). State and local data help establish local program priorities and are necessary for evaluating the impact of suicide prevention strategies.

Nationally, suicide surveillance data come from death certificates. This vital statistics information is available from the National Center for Health Statistics, Centers for Disease Control and Prevention. Medical examiner databases also provide some information related to suicide. The information on rates available from vital statistics databases obviously does not include those deaths misclassified as homicides or accidents, and an unknown number of others misclassified as natural causes, but which may actually be suicides. Information available from death certificates is limited and is not always complete. Prevention efforts would be enhanced by more comprehensive information. However, such information is not now systematically collected.

Data on suicide attempts must come from sources designed for other purposes, such as trauma registries and uniform hospital discharge data sets. Trauma registries provide detailed information about the nature and severity of an injury, the treatment provided, and the status of the patient on discharge from the hospital. However, most trauma registries include only "major" trauma cases, those that require at least a three-day hospital stay (Bonnie et al., 1999). Moreover, many suicide attempts do not lead to traumatic injuries (e.g., overdoses of medicines). Thus, trauma registries have only limited information on suicide attempts.

A uniform hospital discharge data set is another potential source of information on suicide attempts. As suggested by its name, a hospital discharge data set provides information only about those suicide attempts that resulted in hospital treatment. Not all States require either trauma registries or uniform hospital discharge data.

The State of Oregon is unique in that a 1987 law requires hospitals treating a child under the age of 18 for injuries resulting from a suicide attempt to report the attempt to the Oregon Health Division (Hopkins et al., 1995). This data source provides important information for youth suicide prevention programming in Oregon.

Other possible sources of data on suicide attempts include mental health agencies, psychiatric hospitals, poison control centers, universities and colleges, child death review team reports, emergency departments, and surveys. Limitations exist for all of these data sources, such as lack of detail on the circumstances surrounding the suicide attempt (Children's Safety Network, 2000). Detailed information is important because it may lead to increased knowledge of how suicides can be prevented in the future.

One problem in studying nonlethal suicidal behavior is lack of consensus about what terms should be used to refer to these types of behaviors (O'Carroll et al., 1996). Self-injurious behavior exists on a continuum from actions without conscious intent to die, to actions with lethal intent that do not result in death (suicide attempts), to intentional, self-inflicted death (suicide). Often it is difficult to identify a person's intent to die and thus difficult to differentiate attempted suicide from self-injurious behavior in which there was no intent to die.

The objectives established for this goal are designed to enhance the quality and quantity of data available at national, state and local levels on suicide and attempted suicide and ensure that the data are useful for prevention purposes.

Death scene investigations can reveal important information about the circumstances of a suicide and its method. This information can be used to improve understanding of suicide and enhance prevention efforts. Emergency medical technicians, police, medical examiners, and coroners may all contribute to the collection of data.

The detail and specificity of death scene investigations vary by the training and orientation of those who participate in them. The medical examiner or coroner is seldom the first to arrive at the scene of a suicide; prehospital care providers and fire fighters are more likely to arrive first. The core training curricula for such first responders include little attention to death scene investigation in general or suicide scene preservation specifically. In many jurisdictions, prehospital care providers are required to attempt resuscitative efforts even in the face of overwhelming evidence of death. Such resuscitation and transport efforts disrupt evidence. Protocols defining situations that do not require resuscitation and that specify important evidentiary findings that should be preserved would help to address these problems. While law enforcement officers are trained in death scene investigation, that training is primarily geared towards the confirmation or denial of foul play. Additional training in gathering evidence from a suicide scene would provide data that should assist in prevention activities (MacKay, 1997). For example, recording the names of witnesses to the death or close personal contacts of the decedent could help uncover essential information about the decedent, e.g., statements and other behaviors prior to the death, as well as personal and family histories. Testimonies of these individuals may reveal insight into the intent, as well as relevant risk and protective factors.

Follow-back studies consist of the collection of detailed information about the victim, his or her circumstances, the immediate antecedents of the suicide, and other important but less immediate antecedents. Data sources include personal interviews and medical records. Follow-back studies can be used to increase understanding of the causes of suicide and to refine prevention strategies (Berman 1993; Clark & Horton-Deutsch, 1992; Conwell et al., 1996). In some States, child death review teams analyze suicides of young people, and information from these reviews is used to assist in prevention programming.

Consistent use of external cause of injury codes in hospital discharge data and emergency department records would provide an extremely valuable resource for the study and prevention of suicide. Emergency physicians, in particular, have a key role in ensuring that these data are collected. Injury codes include information about the causes and circumstances of injuries (the "how") and, in combination with other information in the medical record, the effect of different injuries on the body. The codes can also be used to obtain information on cost of treatment. External cause of injury codes were developed by the World Health Organization as a part of the International Classification of Diseases. They are standardized internationally and allow consistent comparisons of data among communities, States, and countries (or across time for purposes of evaluation studies). As of January 1997, 17 States had some type of requirement for such coding (Arizona, California, Connecticut, Delaware, Georgia, Maryland, Massachusetts, Missouri, Nebraska, New Jersey, New York, Pennsylvania, Rhode Island, South Carolina, Vermont, Washington, and Wisconsin) (Education Development Center, 1999).

A detailed, Federally-supported data system exists to assist policy-making for motor vehicle related deaths, but such a data system does not exist for violent deaths, including those from suicide. As a result, much of the public debate about violent deaths is based on limited rather than comprehensive data. A national reporting system, which might consist of information derived from a combination of sources, including death certificates, medical examiner or coroner's offices, and law enforcement agencies, would fill a gap in current knowledge by providing consistent, comparable data from all States.

An annual State report on suicide describes the magnitude of the suicide problem in the State and how suicide differentially affects special populations; thus, the data can be used to identify priorities for planning and programming. The data can also help track trends in the suicide rate over time and identify new problems related to suicide, such as changes in the methods for completing suicide or in the suicide rate among certain groups. And finally, the data can help the State evaluate its suicide prevention efforts.

It is estimated that a far greater number of people attempt suicide than is reflected in statistics based on medical care. Some people are treated in a physician's office from which no reporting is ever made, and others are not treated by medical personnel at all. Studies examining nonfatal suicidal behavior have found that over 70 percent of persons attempting suicide never seek health services (Crosby, Cheltenham, & Sacks, 1999). Even the best of existing data on suicide underestimates the burden it places on society (Rosenberg et al., 1987). The Youth Risk Behavior Survey provides one source of information on youth suicide, but there are no other nationally administered surveys that regularly include questions about suicidal behavior. Such information would be quite useful in understanding the true scope of the problem and in designing interventions.

The utility of data can be enhanced significantly when data sets are linked. Such linked data can provide much more comprehensive information about an event, its circumstances, the occurrence and severity of injury, the type and cost of treatment received, and the outcome in terms of both mortality and morbidity (Bonnie et al., 1999). The National Highway Traffic Safety Administration, U.S. Department of Transportation has fostered and supported the development of linked data systems in improving knowledge related to traffic injuries. It has found that many important questions can be answered through analyses based on data linkage. However, significant barriers exist with respect to data linkage, including difficulties in obtaining access to various data sets, high costs, limited resources for developing and maintaining databases, technical difficulties, and issues of confidentiality (Bonnie et al., 1999). Some of these problems are being addressed, however, such as the development of "probabilistic linkage" software that can track individual cases through multiple data sets even in the absence of common identifiers. In short, linking data sets is not a trivial undertaking, but the rewards are often substantial in terms of higher quality data and more complete information.

NSSP Homepage | Table of Contents | Previous | Next

Substance Abuse and Mental Health Services Administration

Centers for Disease Control and Prevention

National Institutes of Health

Health Resources and Services Administration