Annual Review of Public Health - Volume 16, 1995

This article reviews the Bayesian statistical approach to the design and analysis of research studies in the health sciences. The central idea of the Bayesian method is the use of study data to update the state of knowledge about a quantity of interest. In study design, the Bayesian approach explicitly incorporates expressions for the loss resulting from an incorrect decision at the end of the study. The Bayesian method also provides a flexible framework for the monitoring of sequential clinical trials. We present several examples of Bayesian methods in practice including a study of disease progression in AIDS, a comparison of two therapies in a clinical trial, and a case-control study investigating the link between dietary factors and breast cancer.

Exposure assessment is the weakest element in nutritional epidemiologic studies. In the absence of an adequate arsenal of biomarkers of intake in the United States, food frequency questionnaires are widely used to assess habitual frequency of consumption of foods. These tools need to be designed for the population under study, based on prior information on the eating behavior of the population. The questions to be addressed to insure appropriate application of these tools are presented. The influence of various sources and types of measurement error on various scientific hypotheses is addressed. In assessment of nutrient adequacy, information on intra- to interindividual variation of the nutrient or substance of interest is essential. Risk assessment requires examination of sources and extent of bias and differential and nondifferential measurement error within the study. The theory required for error correction is well developed, but rarely carried out because of lack of software and lack of information needed to calibrate the measures.

An ecologic study focuses on the comparison of groups, rather than individuals; thus, individual-level data are missing on the joint distribution, of variables within groups. Variables in an ecologic analysis may be aggregate measures, environmental measures, or global measures. The purpose of an ecologic analysis may be to make biologic inferences about effects on individual risks or to make ecologic inferences about effects on group rates. Ecologic study designs may be classified on two dimensions: (a) whether the primary group is measured (exploratory vs analytic study); and (b) whether subjects grouped by place (multiple-group study), by time (time-trend study), or place and time (mixed study). Despite several practical advantages of ecologic studies, there are many methodologic problems that severely limit causal inference, including ecologic and cross-level bias, problems of confounder control, within-group misclassification, lack of adequate data, temporal ambiguity, collinearity, and migration across groups.

Environmental epidemiological research involves the identification of relationships between previous exposures to putative causative agents and subsequent biological effects with study populations. Such relationships are often hard to fully characterize because of difficulties in accurately quantifying exposure, dose, and effect. Biomarkers are indicators, residing in biological systems or samples, of exposure, dose, effect, or susceptibility. Biomarkers of exposure indicate the presence of previous exposure to an environmental agent; a biomarker of dose bears a quantitative relationship to previous exposure or dose; these include exogenous substances, interactive products, or interactions that change the status of the target molecule. Biomarkers of effect indicate the presence and magnitude of a biological response to exposure to an environmental agent; these include endogenous components, or measures of the functional capacity or state of the system. Biomarkers of susceptibility indicate an elevated sensitivity to the effects of an environmental agent; these include the presence or absence of an endogenous component, or abnormal functional responses to an administered challenge. The development of molecular biomarkers for environmental agents is based upon specific knowledge of metabolism, interactive product formation, and general mechanisms of action. The validation of any biomarker-effect link requires parallel experimental animal and human epidemiological studies.

As a subfield of public health, environmental health is concerned with evaluating and ameliorating the effects of people on the environment and the effects of the environment on people. Separating hazards from risks, and characterizing the magnitude, likelihood, and uncertainty of risks is at the heart of environmental health in the 1990s. To this end, a full range of data is needed, including data that characterize the distribution of hazards, the population potentially at risk, and the contact between people and pollution that creates the risk. Several government-sponsored data systems contain information on a range of exposure estimators. The challenge is to develop meaningful, properly validated models to identify public health needs and evaluate public health programs.

There has been significant concern in recent times about the safety of molluscan shellfish for human consumption. Despite extensive efforts to assure a safe supply of molluscan shellfish, the number of cases of disease and death are still great enough to cause concern among the public. The number of cases of illness and death associated with the ingestion of shellfish falls in the lower end of the range of other similar microbial pathogen-related foodborne disease. Disease and deaths due to viruses and naturally occurring bacteria are now of greatest concern because they are the most often cited casusative agents. The greatest risk of disease or death due to shellfish consumption is among the population with underlying health conditions who choose to consume raw shellfish. Control strategies to limit shellfish-borne disease should focus upon disease and death caused by viruses and naturally occurring bacteria among at-risk populations.

Studies suggest that income replacement is low for many workers with serious occupational injuries and illnesses. This review discusses three areas that hold promise for raising benefits to workers while reducing workers' compensation costs to employers: improving safety, containing medical costs, and reducing litigation.

In theory, workers' compensation increases the costs to employers of injuries and so provides incentives to improve safety. Yet, taken as a whole, research does not provide convincing evidence that workers' compensation reduces injury rates. Moreover, unlike safety and health regulation, workers' compensation focuses the attention of employers on individual workers. High costs may lead employers to discourage claims and litigate when claims are filed. Controlling medical costs can reduce workers' compensation costs. Most studies, however, have focused on costs and have not addressed the effectiveness of medical care or patient satisfaction. Research also has shown that workers' compensation systems can reduce the need for litigation. Without litigation, benefits can be delivered more quickly and at lower costs.

The combination of the burgeoning interest in health, health care reform and the advent of the Information Age, represents a challenge and an opportunity for public health. If public health’s effectiveness and profile are to grow, practitioners and researchers will need reliable, timely informati6n with which to make information-driven decisions, better ways to communicate, and improved tools to analyze and present new knowledge.

“Public Health Informatics” (PHI) is the science of applying Information- Age technology to serve the specialized needs of public health. In this paper we define Public Health Informatics, outline specific benefits that may accrue from its widespread application, and discuss why and how an academic discipline of public health informatics should be developed. Finally, we make specific recommendations for actions that government and academia can take to assure that public health professionals have the systems, tools, and training to use PHI to advance the mission of public health.

The resurgence of tuberculosis confronts policy-makers with difficult legal and ethical questions about the proper use of state power and resources to protect public health. This chapter examines the implications of expanded use of invasive or coercive measures—including directly observed therapy, involuntary detention of noncompliant patients, and forced administration of medications—designed to reduce the risk of tuberculosis transmission and to ensure that those with TB are fully treated. These measures focus attention on the limitations of government power and obligation and on the delicate balance between the demands of civil liberty and the demands of public health.

General health status and a broader concept of quality of life are discussed and methods of widely used surveys are reviewed. A consensus regarding the inclusion of measures of physical, mental, social, and role functioning and general health perceptions is noted for comprehensive assessments of health. A schematic of relationships among condition-specific and generic measures is presented along with results expected for objective and subjective measures of physical and mental dimensions of health. Suggestions are offered for the labeling of disease-specific and generic measures and ways to avoid confounding of content. Applications of health surveys in general population monitoring, health policy evaluation, clinical trials of alternative treatments, monitoring and improving of health care outcomes, and in everyday clinical practice are exemplified and discussed. A unified measurement strategy is proposed and arguments in favor of standardizing the content of health surveys across applications are offered.

This review organizes available data to address current epidemiological understanding of injuries from workplace assaults. We describe the incidence and general populations at risk for fatal and nonfatal work-related injury, and assess available information on risk factors and countermeasures considered important in reducing injury occurrence. Overall rates of occurrence of workrelated homicides and nonfatal injuries are estimated, and those factors that appear to be consistent across studies are explored.

Much of what is known about occupations at high risk for assaults and injuries comes from a small number of documents and special studies within subsets of workplaces. Unfortunately, the potency of specific countermeasures for prevention of violence-related injury in work settings continues to be largely unknown. Numerous fundamental questions remain to be answered: Critical is a full assessment of risk, identification of situations and circumstances amenable to intervention, and evaluations to demonstrate effectiveness.

This paper surveys recent health care reform debates and empirical evidence regarding the potential role for risk adjusters in addressing the problem of competitive risk segmentation under capitated financing. We discuss features of health plan markets affecting risk selection, methodological considerations in measuring it, and alternative approaches to financial correction for risk differentials. The appropriate approach to assessing risk differences between health plans depends upon the nature of market risk selection allowed under a given reform scenario. Because per capita costs depend on a health plan's population risk, efficiency, and quality of service, risk adjustment will most strongly promote efficiency in environments with commensurately strong incentives for quality care.