Search Results

This chapter provides some scientific background of risk assessment and sets up some of the institutional questions concerning the appropriate evidentiary standards needed within torts and administrative law to establish the requisite harms or risks of harm. The discussion here focuses on animal bioassays and human epidemiological studies, two aspects of carcinogen risk assessment relied upon in regulatory institutions to ascertain risks to human beings. The best evidence that proves that a substance causes cancer to human beings is provided by well-done epidemiological studies with large samples and sufficient follow-up. However, the chapter begins by considering animal studies, which is the evidence much more frequently relied upon by regulatory agencies, although it is less typically used in tort law and some jurisdictions give it little credence. There are uncertainties in animal bioassays which are large enough such that two different researchers using exactly the same data points from an animal study can come to much different conclusions.

Epidemiological studies have apparently distinct designs but are unified by their common goal to understand the frequency and causes of disease, by their strategy of seeking associations between exposures (potential causes) and outcomes (disease), by their utilization of the survey method, and by their basis in defined populations. This chapter discusses study design and analysis. Topics covered include classifications of study design; case series, cross-sectional study, case-control study, cohort study, trials, and data analysis and interpretation.

This chapter considers the art of reading and interpreting an epidemiological study, as discussed in the context of epidemiological theory, ethics, and practice. Topics discussed include the interrelationship of theory, methods, and application; fundamental influences on health and disease in populations; scope of epidemiology and specialization; and the practice of epidemiology in public health.

This chapter serves as a guide for developing a course in psychiatric epidemiology; that is, epidemiological methods applied to the study of psychiatric disorders rather than the epidemiology of specific psychiatric disorders. Methodological issues particularly relevant for these pathologies are emphasized, leaving the delineation of the substantive content of different mental disorders to the instructor's interests and expertise. This course does not need to assume prior training in psychiatry or psychopathology, and can be oriented to students at different educational levels. It is expected that students enrolled in this type of course have an interest in public health, medicine or psychology.

This chapter considers dietary factors that may be important in the cause and prevention of cancer. Topics covered include general approaches to the study of diet and cancer, measurement of diet in epidemiological studies, methodological issues in nutritional epidemiology, associations of diet with specific cancers, and prevention of cancer by dietary means.

This chapter traces the development of the epidemiology of cardiovascular disease (CVD). Formal epidemiological research in cardiovascular diseases took several forms, beginning in 1948. The prototype and icon of international investigations was the Framingham Heart Study. The explosion of prospective studies from 1948 onwards, the modern era of research from 1972 to the present, and conceptual evolution in CVD epidemiology are discussed.

This chapter traces the development of the epidemiology of mental disorders. Topics discussed include early epidemiology from 1901–60, improving diagnosis and counting the prevalence of a broad spectrum of disorders, identifying the target population, the scope of psychiatric epidemiological studies, multiple diagnoses and comorbidity, and the current status of epidemiological studies of mental disorders.

Chapter 6 considers the form and spread of spirit concepts, drawing heavily on ethnographic data and emerging theories within epidemiological studies of culture. This chapter suggests why spirit concepts are an ubiquitous feature of human culture, arguing (on the basis of cutting‐edge research in the cognitive sciences) that such concepts have a distinct cognitive advantage in cultural transmission.

The cost, complexity, and risk of selection bias often associated with primary data collection has led to the use of disease registries and databases as an alternative data source for studies of many epidemiological and clinical questions. Medical registries and administrative databases are attractive options. They often contain information on large study populations, and they can provide information on exposures, disease outcomes, and some potential confounding factors for subjects in those populations. Thus medical databases can be very useful for the efficient study of etiologic associations in large populations, and for the evaluation of utilization, effectiveness and safety of medical interventions in clinical settings. This chapter presents an approach for teaching use of routine medical databases covering teaching objectives, teaching method and format, and assessing students' achievements.

This chapter introduces the ways in which data are manipulated, with a warning that the end results are often artificial measures that change (and sometimes distort) the perception and interpretation of risk, a matter of utmost concern to a science where communication between researchers, health professionals, and the public is critically important. Topics discussed include proportional morbidity or mortality ratio; relative risk; measurements to assess the impact of a risk factor in groups and populations; presentation and interpretation of epidemiological data in applied settings; and avoidable morbidity and mortality and life-years lost.

This chapter utilized the central epidemiological strategy — comparison of populations — to discuss bias. Bias, in an epidemiological context, arises when errors affect comparison groups unequally. Since this is often the case, perhaps always the case, bias is a central issue in epidemiology. Classification of error and bias and a practical application of the research chronology schema are discussed.

This chapter reviews ways in which chronic pain has been measured in epidemiological studies. It summarizes a simple approach which emphasizes the need for core standard definitions to describe pain presence and persistence, i.e., a discussion of the principles underlying work on a core definition of back pain and how they might be rolled out to epidemiological definitions of chronic pain generally.

This chapter discusses the concept of epidemiology. Topics covered include definition of epidemiology and a statement of its central paradigm; directions in epidemiology and its uses; epidemiology as a science, practice, and craft; the nature of epidemiological variables; and the basic tools of epidemiology.

Health impact assessors who wish to advise policy makers and base their health impact assessment (HIA) on the fullest scientific evidence must understand the potential and pitfalls of epidemiology. This chapter explores how epidemiological reasoning can give insight into the causal mechanisms by which decisions impact on health, goes some way towards estimating the size of those impacts, and thereby provide a basis for prediction.

Biomarkers are increasingly being incorporated into epidemiologic studies of cancer etiology, often referred to as molecular epidemiology. This chapter focuses on the use of biomarkers in the context of etiologic research. It begins with a discussion of the development and characterization of biomarkers for use in epidemiologic studies, with a particular emphasis on understanding the components of variance and their impact on estimates of disease risk. It then discusses key issues in the use of exposure, intermediate end points, genetic susceptibility, and tumor biomarkers; and reviews the main epidemiologic study designs that use biomarkers. The chapter concludes with comments on collecting and processing biologic samples for use in molecular epidemiology studies.

The exposure to any pollutant is determined by the concentration of the pollutant in the environment, its specific character, and the duration and frequency of contact. Estimates of these characteristics can be obtained instrumentally (i.e., using a monitoring device), via questionnaires, through direct observation, or through the use of biomarkers. This chapter focuses on the measurement of personal exposure with instruments or biomarkers. Although personal monitoring is labour-intensive and expensive, it remain a valuable technique because it can dramatically improve exposure assessment in epidemiological studies, thus adding to the scientific credibility of the studies.

This chapter provides a general background to spatial epidemiological studies. It discusses issues that are pertinent to their analysis and interpretation, including the underlying statistical methodology. It considers the aims and use of spatial analyses in epidemiology. It distinguishes between four types of study: disease mapping, geographical correlation studies, risk assessment in relation to a point or line-source, and cluster detection and disease clustering.

This chapter addresses some of the issues affecting the interpretation of spatial epidemiological studies. It separates the two concepts of bias and confounding. In general, bias and confounding are major problems affecting all observational studies, and are not special to spatial studies. However, there are several forms of bias that are specific to spatial studies. These include selection effects, denominator bias, exposure inaccuracy bias and the errors-in-variables problem, spatial dependency, significance tests, and ecological bias. The chapter also discusses socio-economic confounding as this is a major potential source of bias in spatial epidemiology.

This chapter reviews specific problems of the assessment of disease risk in relation to a pre-specified source. The chapter proceeds as follows. Section 9.2 provides a context for point-source studies and describes a general statistical framework. Section 9.3 describes a number of conventional epidemiological techniques while Section 9.4 describes semi-parametric statistical approaches to modelling. Section 9.5 reviews regression methods. Section 9.6 illustrates a number of methods using data on the incidence of stomach cancer in the proximity of a municipal incinerator. Section 9.7 provides a concluding discussion.

This chapter is divided into two parts. The first part describes the key study designs used in clinical and epidemiological studies of causation. The designs are defined by two features: whether the study subjects are selected by their exposure or by their outcome, and the time dimensions of the study. The second part describes the key strengths and weaknesses of each type of study, with many examples. Self test questions are provided at the end of the chapter.