Geographic patterns of fatal abuse or neglect in children younger than 5 years old, United States, 1979 to 1988.

OBJECTIVE: To examine geographic patterns of fatal child abuse or neglect (CAN) among children younger than 5 years old. DESIGN: A death certificate-based model to estimate the occurrence of fatal CAN. SETTING: United States, 1979 to 1988. PARTICIPANTS: The population of children younger than 5 years old. INTERVENTIONS: None. MAIN RESULTS: We estimate that from 868 to 1815 deaths annually occur among children younger than 5 years old from CAN. The lower figure is the estimate of confirmed CAN, and the higher is the estimate of the sum of confirmed, probable, and possible CAN. Death rates were highest in the South and West, intermediate in the North Central, and lowest in the Northeast. A threefold difference was noted between rates in the lowest- and highest-ranking states (ie, Connecticut, 2.9 to 5.2 per 100,000, and Nevada, 6.7 to 15.4 per 100,000, respectively). When the 39 largest metropolitan areas were ranked, a similar variation between the lowest and the highest was observed (ie, Boston, Mass, 2.7 to 5.5 per 100,000, and Phoenix, Ariz, 6.6 to 15.5 per 100,000, respectively). CONCLUSION: Understanding the sizable geographic variation in CAN deaths rates could lead to effective interventions. If the US fatality rate were reduced to that of Connecticut, between 434 and 908 fewer CAN deaths might occur annually.

Public health policy for preventing violence.

The current epidemic of violence in America threatens not only our physical health but also the integrity of basic social institutions such as the family, the communities in which we live, and our health care system. Public health brings a new vision of how Americans can work together to prevent violence. This new vision places emphasis on preventing violence before it occurs, making science integral to identifying effective policies and programs, and integrating the efforts of diverse scientific disciplines, organizations, and communities. A sustained effort at all levels of society will be required to successfully address this complex and deeply rooted problem.

PIP: Violence is a major contribution to premature death, disability, and injury. In America, there is an epidemic of violence, which threatens not only the physical health, but also the integrity of basic social institutions such as the family, the communities, and the health care system of the public. In this paper, the new vision for violence prevention embodied in the public health approach is discussed. It shifts the focus of the society in the way violence is addressed, from reacting to the problem to changing the social, behavioral, and environmental factors that cause violence. The emphasis is on preventing violence before it occurs, making science integral in identifying effective policies and programs, and integrating the efforts of diverse scientific disciplines, organizations, and communities. A sustained and coordinated effort to prevent violence will be necessary at all levels of society to address this complex and deeply rooted problem.

Use of national data systems for firearm-related injury surveillance.

INTRODUCTION: The purpose of this paper is to provide an overview of federal data systems that report national data on fatal and nonfatal firearm-related injuries and associated risk factors and behaviors. RESULTS: There are 13 federal data systems that provide useful information for national surveillance of firearm-related injuries in the United States. Each data system has useful features and limitations. Each provides a different methodologic approach to capture data for monitoring and characterizing firearm-related deaths or injuries, or behavioral risks associated with unintentional and violent events. CONCLUSION: Although much progress has been made over the past decade to improve national data on firearm-related injuries, many gaps still remain. A mechanism is needed to better coordinate and integrate federal efforts to collect, analyze, and disseminate data on firearm-related injury.

Building a foundation for suicide prevention: the contributions of Jack C. Smith.

Among his many other accomplishments, Jack C. Smith nurtured the early development of efforts by the Centers for Disease Control and Prevention (CDC) to address suicide as a public health problem. Smith's vision was to achieve suicide prevention through epidemiology, and his vision shaped the emergence of suicide as a public health issue. With his typical enthusiasm and inherent ability to insinuate himself into critical social networks, Smith spearheaded CDC's initial suicide surveillance activities and established strong partnerships between CDC and the National Institute of Mental Health (NIMH) and the American Association of Suicidology (AAS). These surveillance activities and relationships were the foundation on which subsequent research and programmatic activities addressing suicide as a public health problem were built at CDC. In this paper we document Smith's role in the development of the public health approach to suicide prevention. We also articulate the conceptual basis for a public health approach to suicide and discuss future directions for public health in the prevention of suicide and suicidal behavior. While Smith also made important contributions to development of CDC efforts to address homicide, his special interest was suicide; therefore, this article will emphasize his contributions to this area.

Firearm-related injury surveillance. An overview of progress and the challenges ahead.

Firearm-related injuries pose a serious public health problem in the United States and are increasingly the focus of public health concern. Despite the magnitude of this problem, ongoing and systematic collection of data on firearm-related injuries to help guide research and policy development has been lacking. The further development of firearm-related injury surveillance systems can provide an objective source of information for policy. Beginning in the mid-1980s, the Centers for Disease Control and Prevention's National Center for Injury Prevention and Control began to support the development of firearm-related injury surveillance systems by augmenting existing national- and state-level data collection systems and establishing cooperative agreements with state and local health departments to identify optimal firearm-related injury surveillance practices. Some progress has been made in improving the capacity to undertake firearm injury surveillance at national, state, and local levels for mortality, morbidity (including disability), and risk/protective factors, but much work remains to be done. The development of state and local firearm-related injury surveillance systems provides the clearest potential for linking basic information on firearm-related injuries to action, given the critical role that states have in both public health surveillance and regulation of firearms. Broader application of external cause-of-injury codes, increased standardization and validation of definitions and data-collection instruments, improved methods for identifying firearm characteristics and types, and the identification of efficient techniques for linking health and criminal justice data sources are among the key challenges we face as we try to build a more uniform system for monitoring firearm-related injuries in the United States.

Prevention of youth violence: rationale and characteristics of 15 evaluation projects.

Interpersonal violence is a major cause of injury, disability, and death, especially among youth. Evaluations of 15 youth violence-prevention projects are under way. Public health is concerned about health problems that need to be addressed via collective action. Public health involvement in addressing interpersonal violence among youths brings an emphasis on primary prevention, a systematic and scientific process, and integrative leadership. Few quantitative evaluations of violence-prevention projects have been done. The interventions are scientifically based and use a spectrum of strategies. Individually oriented strategies are more common than those directed toward peers, families, schools, or communities. Each project has a rigorous evaluation design. Twelve are randomized. Sample sizes range from 180 to 10,000. Participants range in age from 5 to 18 years, although most are in the middle-school years (11-14 years). At baseline, intervention and comparison groups are similar. Baseline data demonstrate high frequency of violent behavior, weapon carrying, and exposure to violence among the youthful participants. Field intervention and evaluation research is difficult and expensive. Difficulties encompass organizational programatic, and scientific issues; these difficulties reduce scientific interest and financial support for projects such as these. Public health has an important role to play in reducing violence. These projects will make important contributions to that task.

Trends in fatal firearm-related injuries, United States, 1962-1993.

OBJECTIVE: Our objective was to review historical trends in U.S. fatal firearm-related injuries for the years 1962-1993. METHODS: Using mortality data from the National Center for Health Statistics and population estimates projected from census data, we calculated national age-adjusted mortality rates and examined trends over the 32-year period. Data were also examined by type of firearm-related death (unintentional, suicide, homicide, legal intervention, and undetermined intention), race, gender, and age group. RESULTS: During the 32-year period, the total number of firearm-related deaths increased by 137%, from 16,720 in 1962 to 39,595 in 1993. Suicide and homicide were responsible for most firearm fatalities. Rates for both firearm suicides and firearm homicides increased over time, while rates for unintentional, legal intervention, and undetermined intention decreased. The highest rates and widest variation in total firearm-related mortality occurred among African-American men (35.2/100,000 to 84.5/100,000). Persons 15-19, 20-24, and > or = 75 years of age experienced the largest changes in rates during recent years; total firearm mortality was higher for the younger age groups (15-19, 20-24) during 1990 through 1993 than any other time during the 32-year period. CONCLUSIONS: These surveillance data help characterize trends over time and the magnitude of firearm-related mortality and identify groups at risk. However, further efforts to improve our understanding of firearm-related deaths and injuries, such as expansion of current surveillance to include information about morbidity associated with firearms and additional epidemiologic research to identify modifiable individual and societal risk factors, are necessary.

Fighting as a marker for multiple problem behaviors in adolescents.

PURPOSE: Behaviors that put adolescents at risk frequently occur together. To help identify high-risk adolescents, we analyzed a national, self-reported behavior survey of high school students to assess the suitability of fighting as a marker for students with multiple problem behaviors. METHODS: A cross-sectional cluster survey of 11,631 U.S. high school students in 1990 was used to compare the prevalence of recent problem behaviors among all students and those who fight. RESULTS: One (8%) of every 12 students was in a fight during the 30 days before the survey. Reported problem behaviors were prevalent among fighters: during the previous 12 months, 24% attempted suicide; during the previous 30 days, 26% carried a firearm, 13% used cocaine, and 39% drove a motor vehicle while intoxicated; during the previous 3 months 41% had two or more sex partners; and 45% had sexual intercourse and did not use a condom the last time they had sex. Of all students, fighters accounted for 22% of those who reported attempting suicide, 49% carrying a firearm, 46% using cocaine, 18% driving while intoxicated, 25% having sex with multiple partners, and 11% not using condoms. Three or more of these six problem behaviors were reported by 26% of the fighters. The problem behaviors were all positively correlated, and the first principal component accounted for 35% of the total variation among the individual variables.

Suicide and homicide among Hispanics in the southwest.

A study of suicide and homicide among Hispanics of Mexican origin (Mexican Americans) focused on five southwestern States--Arizona, California, Colorado, New Mexico, and Texas--where more than 60 percent of all Hispanics in the United States reside. And 85 percent of them are Mexican Americans. Data were obtained on all suicides and homicides among Hispanics and Anglos (white non-Hispanics), using Anglos as a comparison group. Results for suicide showed the suicide rate for Hispanics (9.0 per 100,000) to be less than the national rate for whites (13.2) and half that of the Anglos residing in the same area (19.2). The lower suicide rate for Hispanics relative to Anglos is seen for both males and females. For homicide, the overall rate for Hispanics (20.5) was more than 2 1/2 times that of Anglos (7.9). The rate for Hispanic men (39.3) was more than three times the rate for Anglo men (11.4).

Comparing death certificate data with FBI crime reporting statistics on U.S. homicides.

Both the National Center for Health Statistics (NCHS) Mortality System and the Federal Bureau of Investigation (FBI) Uniform Crime Reporting System measure the numbers and characteristics of homicide in the United States. There are important differences, however, in both the substance and the quality of the information that the two systems collect. The NCHS mortality system reported an average of 9 percent more homicides nationally than did the FBI crime reporting system during the 1976-82 period. Variations did occur in the average ratios of the frequencies of homicides reported by the two systems across age, race, and sex subgroups and geographic areas. The major source of the ascertainment difference between the NCHS and the FBI systems is thought to be incomplete voluntary reporting to the FBI by participating law enforcement agencies and lack of reporting by nonparticipating agencies. The proportions of homicides among corresponding demographic categories in the two systems is remarkably similar despite the difference in ascertainment. This congruence of the distributions of reported homicides supports the idea that inferences drawn from analysis of variables in one of these systems will be valid for the population reported on by the other system.

The emergency department in surveillance of attempted suicide: findings and methodologic considerations.

The authors conducted one of the first active, population-based public health surveillance systems for detecting suicide attempts in the United States. Surveillance was conducted in all four hospital emergency departments serving a county suburban to Atlanta, GA, with a population of 426,000. Emergency department staff gathered information from all patients who presented with an intentionally self-inflicted injury (suicide attempt) or with thoughts about self-injury (suicidal ideation). During an 18-month period in 1988 and 1989, 798 suicide attempt-related patients were reported, for a rate of 124.7 per 100,000 county residents per year. Females had a higher attempted suicide rate than males, but males had a higher completed suicide rate. Ingestion of drugs or poison was the most common method of attempted suicide (71.1 percent), and use of firearms was the most common method of completed suicide (69.8 percent). In comparing reported cases with those found by reviewing emergency department log books, the authors found that the case reports were 58 percent complete and that surveillance reporting was highly representative of all cases requiring emergency transport. The authors conclude that emergency department-based surveillance for attempted suicide is feasible. It can provide representative data that may be used to monitor trends in attempted suicide and to define high-risk groups. Such surveillance may also allow timely detection of suicide attempt clusters, facilitating prompt intervention.

Comparison of suicides among Anglos and Hispanics in five southwestern states.

Little is known about deaths resulting from self-inflicted violent behavior for Hispanic Americans. To learn more about suicide for Hispanics of Mexican origin (Mexican Americans) we focused on the five southwestern states (Arizona, California, Colorado, New Mexico, and Texas) in which more than 60% of all Hispanics in the United States reside (86% of whom are Mexican American). We obtained data on the number of suicide deaths in the white population, with Hispanics and Anglos (white, non-Hispanic) identified separately. Suicides of Anglos were used as a comparison group. Results show that the suicide rate for whites in the five southwestern states (15.6) was almost one-fourth higher than the rate for whites nationally (12.7). Suicide rates for the two ethnic groups, however, showed the rate for Hispanics (9.0) to be less than the national rates for whites (12.7) and one-half that of Anglos residing in the same area (17.3). The lower suicide rate for Hispanics relative to Anglos is seen for both males and females. The ratio of male and female suicides for Hispanics was almost twice that for Anglos (4.0 to 1 for Hispanics and 2.3 to 1 for Anglos). More than 1 in 3 Hispanic men who committed suicide and more than 1 in 4 Hispanic women who committed suicide are under the age of 25.

Alcohol consumption and nearly lethal suicide attempts.

We conducted a case-control study of the association between nearly lethal suicide attempts and facets of alcohol consumption; namely, drinking frequency, drinking quantity, binge drinking, alcoholism, drinking within 3 hours of suicide attempt, and age began drinking. Subjects were 13-34 years of age. In bivariable analyses, all measures were associated with nearly lethal suicide attempts. Odds ratios ranged from 2.4 for alcoholism to 7.0 for drinking within 3 hours of attempt. All exposure variables except age began drinking exhibited a J-shaped relationship between alcohol exposure and nearly lethal suicide attempt. After controlling for potential confounders and other measures of alcohol exposure, drinking within 3 hours of attempt remained most strongly (odds ratios > 6) associated. Alcoholism remained significantly associated in most models, but at lower strength.

Medical conditions and nearly lethal suicide attempts.

Physical illness has been studied as a risk factor for suicidal behavior, but little is known about this relationship among younger persons. We conducted a population-based, case-control study in Houston, Texas, from November 1992 through September 1995. The final sample consisted of 153 case- and 513 control-subjects aged 13 to 34 years. Case patients were identified at hospital emergency departments and met criteria for a nearly lethal suicide attempt. Control subjects were recruited via a random-digit-dial telephone survey. Case patients were more likely than controls to report having any serious medical conditions (crude OR = 3.23; 95% CI = 2.12-4.91). After controlling for age, race/ethnicity, alcoholism, depression, and hopelessness, the adjusted odds ratio for men was 4.76 (95% CI = 1.87-12.17), whereas the adjusted odds ratio for women was 1.60 (95% CI-0.62-4.17), suggesting that young men with medical conditions are at increased risk for nearly lethal suicide attempts. Increased efforts to identify and appropriately refer these patients are needed.

An unmatched case-control study of nearly lethal suicide attempts in Houston, Texas: research methods and measurements.

This article details the research methods and measurements used in conducting a population-based, case-control study of nearly lethal suicide attempts among persons aged 13-34 years, residing in Houston, Texas. From November 1992 to July 1995, we interviewed 153 case subjects presenting at one of three participating hospital emergency departments and used random digit dialing to identify 513 control subjects residing in the same catchment area in which cases were enlisted. Unlike most research in this area, this study was designed to extend our understanding of suicidal behavior and prevention activities beyond identification and treatment of depression and other mental illnesses. We discuss the overall strengths and weaknesses of our study design and conclude that this methodology is well suited for studying rare outcomes such as nearly lethal suicide.

The influence of geographic mobility on nearly lethal suicide attempts.

Teenagers and young adults are very mobile and mobility has been identified as a potential risk factor for suicidal behavior. We conducted a population-based, case-control study of nearly lethal suicide attempts with 153 cases and 513 controls. Study participants were asked about changing residence over the past 12 months. Results indicate that moving in the past 12 months is positively associated with a nearly lethal suicide attempt (adjusted odds ratio of 2.1, with 95% confidence interval of 1.4-3.3), as are specific characteristics of the move (e.g., frequency, recency, distance, and difficulty staying in touch). These findings confirm and extend prior ecologic research by demonstrating a relationship, at the individual level, between the geographic mobility of adolescents and young adults and nearly lethal suicide attempts.

New directions in violence prediction: the public health arena.

Injury resulting from interpersonal violence is now recognized as an important public health problem. Consequently, the public health community is applying its perspectives and strategies to the goal of preventing violence. The public health approach emphasizes preventing injuries due to interpersonal violence from occurring or recurring, rather than treating the health consequences of these injuries. Four interrelated steps are used to develop information to guide the development of prevention strategies: public health surveillance, risk group identification, risk factor exploration, and program implementation/evaluation. The ability to predict which people are at greatest risk of injury (or producing injury) and the relative effectiveness and costs of alternative prevention strategies are central to decisions influencing the nature and focus of public health prevention strategies. In order to improve the information base on which decisions concerning violence prevention strategies are founded, the following activities should be given priority: (a) developing surveillance systems for morbidity associated with interpersonal violence; (b) precisely identifying risk groups for nonfatal violent events; (c) applying case-control methods to the exploration of potentially modifiable risk factors for injuries and violent behaviors; and (d) rigorously evaluating extant programs that are intended to prevent interpersonal violence or modify a suspected risk factor for violence.

Regional variation in homicide rates: why is the West so violent?

Scientists have long been interested in the fact that the South has consistently had the highest crude homicide rates in the United States. Past investigations, however, have generally been predicated on the assumption that this geographic pattern was not attributable to or substantially altered by the age or race structures of the populations being compared. In this study, we calculated age-adjusted homicide rates for each of three race categories--white, black, and other--for each state and region in the United States in 1980. We found that for each race group, homicide rates were highest, not in the South, but in the West. Moreover, homicide rates for blacks were lower in the South than in any other region of the country. We infer that, for 1980 at least, the high crude homicide rate in the South results from the mutual effect of two factors: (1) blacks have very high homicide rates compared with whites, and (2) blacks make up a larger proportion of the population in the South than in other regions of the country. It remains to be determined whether the age-adjusted, race-stratified rates of past decades also show this pattern.

Mortality associated with the use of upper-body control holds by police.

A review of police records for the nine-year period from 1974 to 1982 identified 20 men who died following restraint by an upper-body control hold while in the custody of a large, urban police department. Using evidence from autopsy findings and police reports of events immediately preceding death, we concluded that control-hold use was associated with death in 19 of the 20 cases. This investigation points to three factors potentially associated with control hold-related death that deserve further investigation: Phencyclidine (PCP) use, sickle cell trait, and stress-related arrhythmias in the heart. PCP was detected in blood or other tissues from 6 of 17 decedents tested. Intravascular red blood cell sickling was found at autopsy in 4 of 14 black decedents (29%). Four decedents had some indication of cardiovascular abnormalities.