Annual Review of Public Health - Volume 19, 1998

We describe an econometric technique, instrumental variables, that can be useful in estimating the effectiveness of clinical treatments in situations when a controlled trial has not or cannot be done. This technique relies upon the existence of one or more variables that induce substantial variation in the treatment variable but have no direct effect on the outcome variable of interest. We illustrate the use of the technique with an application to aggressive treatment of acute myocardial infarction in the elderly.

Assessing occupational exposures in community-based studies is a challenge for investigators because there are no standardized or validated approaches for collecting information regarding occupational history. The strengths and limitations of the methods available for assessing occupational exposures are reviewed. In community-based case-control studies, the prevalence of most chemical agents is low. The common sources of misclassification in these studies are addressed, as are strategies for dealing with misclassification bias. Methods to assess the presence and magnitude of differential reporting by cases and controls are outlined, together with analytic strategies to improve the classification of occupational exposures.

Multiple observational studies suggest a marked reduction in risk of coronary heart disease (CHD) associated with postmenopausal estrogen use. A new meta-analysis presented here extends these results to estrogen plus progestin regimens. Although the findings from observational studies are strong and consistent, and there are several plausible mechanisms by which estrogen might reduce risk for CHD, most of the known biases would tend to exaggerate estrogen's benefit. Further, estrogen therapy clearly increases risk for endometrial hyperplasia and cancer, venous thromboembolic events and gallbladder disease, and long-term use probably also increases the risk of breast cancer. Therefore, until findings from randomized trials confirm and quantitate the benefit of estrogen therapy for prevention of CHD, we believe it should not be recommended to all postmenopausal women.

Beta-carotene and retinoids were the most promising agents against common cancers when the National Cancer Institute mounted a substantial program of population-based trials in the early 1980s. Both major lung cancer chemoprevention trials not only showed no benefit, but had significant increases in lung cancer incidence and in cardiovascular and total mortality. A new generation of laboratory research has been stimulated.

Rational public health recommendations at this time include: 1. Five-A-Day servings of fruits and vegetables, a doubling of current mean intake; 2. systematic investigation of the covariates of extremes of fruit and vegetable intake; 3. discouragement of beta-carotene supplement use, due to adverse effects in smokers and no evidence of benefit in non-smokers; 4. multilevel research to develop and evaluate candidate chemoprevention agents to prevent lung and other common cancers; and 5. continued priority for smoking prevention, smoking cessation, and avoidance of known carcinogens in the environment.

The increasing incidence of breast cancer in the United States and the international variation in risk have led to speculation that environmental risk factors are an important cause of breast cancer. We review the epidemiologic evidence on the breast cancer risk associated with ambient environmental exposures experienced passively by the US population, and discuss the difficulties associated with measurement of specific exposures in environmental studies. We review geographic variation of breast cancer rates in the United States, and exposure to organochlorines, ionizing and electromagnetic radiation, and passive smoking. Results are inconclusive but do not support a major role of environmental risk factors in the etiology of breast cancer.

Cost-effectiveness analysis, an analytic tool that expresses as a ratio the cost of obtaining an additional unit of health outcome, can help decision makers achieve more health protection for the same or less cost. We characterize the state of the cost-effectiveness analysis literature by reviewing how this technique is applied to various clinical and public health interventions. We describe the results of cost-effectiveness analyses for over 40 interventions to reduce cancer, heart disease, trauma, and infectious disease. The cost-effectiveness ratios for these interventions vary enormously, from interventions that save money to those that cost more than $1 million per year of life gained. The methods used to derive the cost-effectiveness ratios also vary considerably, and we summarize this variation within each health area. Greater uniformity of analytical practice will be necessary if cost-effectiveness analysis is to become a more influential tool in debates about resource allocation.

The central and eastern European countries that composed the former Eastern Bloc have experienced an alarming decline in public health since the dissolution of the Soviet Union. Death rates have increased in most age groups. Life expectancy, especially among males, has decreased in many countries; in Russia, male life expectancy dropped by six years between 1989 and 1994. By 2020, these countries are projected to have smaller increases in life expectancy than any other geographic region. The conditions responsible for the excess mortality are cardiovascular disease, cancer, and injuries among adults. The major factors in the sharp increase are poverty, social disintegration, and crime, overlaid on historically high rates of smoking, alcohol use, and psychosocial stress. Environmental pollution, although common and sometimes severe in the former Eastern Bloc, is another, albeit not the chief, cause of the sharp decline in public health since 1989.

Community-based research in public health focuses on social, structural, and physical environmental inequities through active involvement of community members, organizational representatives, and researchers in all aspects of the research process. Partners contribute their expertise to enhance understanding of a given phenomenon and to integrate the knowledge gained with action to benefit the community involved. This review provides a synthesis of key principles of community-based research, examines its place within the context of different scientific paradigms, discusses rationales for its use, and explores major challenges and facilitating factors and their implications for conducting effective community-based research aimed at improving the public's health.

The late 1980s marked the beginning of a sea change in research and intervention on tobacco use by young people, when the first studies were published demonstrating the ease with which adolescents could purchase cigarettes. Since then concern with the commercial availability of tobacco products to youth has grown, becoming deeply entrenched in both research and policy efforts. This paper reviews the history of restrictions on youth access to tobacco, the evidence that commercial sources of tobacco are available and important to adolescents, and the rationale for various components of policies to restrict access. We also review the recent expansion of policies addressing youth access to tobacco at all levels of government, the tobacco industry response to this expansion, and issues that have arisen as this expansion has developed.

The epidemiologic concept of tick-transmitted diseases has increased in importance with the recognition of the emerging infectious diseases, Lyme borreliosis, human monocytotropic and granulocytotropic ehrlichioses, and three different babesioses. Effective public health control of these diseases would depend upon critical knowledge of the vector biology of the ticks that transmit them. Rocky Mountain spotted fever and the human ehrlichioses are life-threatening yet treatable diseases. A major problem remains establishment of the diagnosis when treatment decisions are being made. Clinical manifestations, other than erythema migrans for Lyme borreliosis, do not provide strong diagnostic clues. Ehrlichiae or babesiae are often not detected in peripheral blood smears. Frequently there are no antibodies to these diverse agents at the time of presentation, and isolation does not yield sensitive and timely results. Polymerase chain reaction, still a research tool, promises the greatest sensitivity, specificity, and timeliness. Prevention by vaccines is not yet a reality, although OspA-based vaccines offer hope for the prevention of Lyme disease.

Between 1985 and 1992, serious youth violence in the United States surged to unprecedented levels. The growing use of firearms to settle disputes has contributed to this phenomenon. Youth are most often victimized by one of their peers. In response to this problem, a wide variety of programs have been implemented in an attempt to prevent youth violence or reduce its severity. Few have been adequately evaluated. In general, interventions applied between the prenatal period and age 6 appear to be more effective than interventions initiated in later childhood or adolescence. Community-based programs that target certain high-risk behaviors may be beneficial as well. A sustained commitment to evaluation research is needed to identify the most effective approaches to youth violence prevention.

Efforts to reduce the toll of bicycle-related head injuries illustrate how the basic public health principles of surveillance, epidemiologic study, intervention, and evaluation can have a substantial impact on an injury problem, using a variety of injury-prevention strategies. Head injuries are the leading cause of serious morbidity and mortality from bicycle crashes. Helmets have been shown to reduce bicycle-related head injuries for cyclists of all ages involved in all types of crashes including those with motor vehicles. Helmet use has been promoted using educational campaigns, helmet subsidies, and legislation. Careful evaluation of these strategies has shown that these interventions increase helmet use and decrease the incidence of bicycle injuries. The model developed for the prevention of bicycle injuries is widely applicable to other injury problems.

Important behavioral sources of risk for cancer have been identified and programs to reduce this risk have been initiated. The most heavily studied behavioral risk is tobacco use, but considerable attention has also gone to prevention of skin cancer by modifying sun exposure. Research has shown a strong link between sun exposure and skin cancer, and opportunities to prevent this cancer by changing behavior should be effective. Unlike unmodifiable sources of risk, sun exposure should be modifiable through behavioral intervention. This review considers research in this expanding area, concentrating on predictors of sun protective behaviors, modification of sun exposure, and surveillance for early signs of cancer. The elements of behavior change programs that contribute to successful intervention as well as barriers to successful behavior change and future directions for this effort are considered.

Cigarette smoking continues to place an enormous burden on US health systems. The treatment of tobacco-dependent persons is vital to the public health, since unmitigated smoking could lead to the development of disease and premature death in as many as 20 million current smokers in the United States. Much of this tobacco-related morbidity and mortality could be prevented by the application of existing treatments, as well as newer ones. Access to safe and effective treatments was significantly expanded in 1996, when two forms of nicotine replacement therapy (NRT), the nicotine gum and patch, became available for over-the-counter sale. Expanded access has led to significantly increased utilization and quitting. Even with these advances, however, the scope of this tobacco dependence demands that we explore alternative treatment strategies, including exposure reduction therapy (ERT) as a possible option for smokers not currently making quit attempts. In this review, we present principles to guide the development of such a therapy, and for its application, should it be proven effective.

The purpose of this review is to update research on the prevention of alcohol-related traffic deaths since the 1988 Surgeon General's Workshop on Drunk Driving. Four primary areas of research are reviewed here: (a) general deterrence policies, (b) alcohol control policies, (c) mass communications campaigns, including advertising restrictions, and (d) community traffic safety programs. Modern efforts to combat drunk driving in the United States began with specific deterrence strategies to punish convicted drunk drivers, and then evolved to include general deterrence strategies that were targeted to the population as a whole. Efforts next expanded to include the alcohol side of the problem, with measures installed to decrease underage drinking and excessive alcohol consumption. In the next several years, greater efforts are needed on all these fronts. Also needed, however, are programs that integrate drunk driving prevention with other traffic safety initiatives.

This paper examines the results of population-level interventions conducted in three settings: entire communities, worksites, and schools. Four major conclusions are discussed: (a) Directions for the next generation of community-based interventions include targeting multiple levels of influence; addressing social inequalities in disease risk; involving communities in program planning and implementation; incorporating approaches for “tailoring” interventions; and utilizing rigorous process evaluation. (b) In addition to randomized controlled trials, it is time to use the full range of research phases available, from hypothesis generation and methods development to dissemination research. (c) The public health research agenda may have contributed to observed secular trends by placing behavioral risk factors on the social and media agendas. (d) The magnitude of the results of community intervention trials must be judged according to their potential public health or population-level effects. Small changes at the individual level may result in large benefits at the population level.

This paper outlines a conceptual framework of organizational diversification and assesses the state of empirical research on hospital organizational change. The literature on economic organization of hospitals, one of the most developed branches of health services research, still has only weak ties to economic theory. Evolving physician-hospital organizations do not fit into existing frameworks based on horizontal integration, vertical integration, or diversification. Empirical research has primarily focused on horizontal integration, and cause-effect relationships are often obscured by models that depart from economic theory and lack controls for self-selection bias. Recent empirical studies indicate that hospital mergers had moderate, rather than dramatic, effects on the rate of change in operating costs, staffing, and scale. Mergers rarely resulted in hospital closure, but were as likely to result in acute care consolidation and restructuring as in conversion to non-acute inpatient uses. While administrative costs were higher in for-profit than non-profit system hospitals, total costs were similar. System hospitals had lower marginal and average costs per stay than independent hospitals. Hospital vertical integration into subacute care was largely an artifact of the governmental uniform pricing system, which encouraged vertical integration. Hospitals that shared governance or financial risks with physicians outperformed those with high levels of physician governance and financial integration (e.g. stock ownership). Formal physician-hospital organizational arrangements often served to coordinate managed care contracting or to forge links with primary care group practices. Hospital diversification into related services improved short-term financial performance over unrelated diversification, although long-term performance was similar.

Medicare offers nearly universal, but limited, coverage for the elderly. The vast majority of beneficiaries therefore obtain supplemental coverage, or they enroll in HMOs to gain extra benefits at substantially lower or zero cost. This is possible because of reduced utilization and costs, as well as favorable selection of lower-risk enrollees into HMOs. Competition from HMOs may lower local fee-for-service costs as well. Quality and satisfaction measures are quite balanced, with some results showing better HMO performance and some worse. The absence of adequate risk-adjusted payments to HMOs, however, gives them little incentive to develop high-quality programs for the sickest enrollees.