Epidemiologic methodology in public health: new frontiers & challenges

Abstract

The paper describes new and innovative applications of the epidemiologic methodology outside the traditional domain of diseases, infectious and noninfectious.Epidemiology, being a methodological discipline, has no coherent subject matter but finds applications in various disciplines. It was used in the study of epidemics of infectious diseases even before the term epidemiology was coined. Over the past 50 years the epidemiologic methodology has been extended to the study of chronic non-communicable disease. The epidemiologic methodology has been extended, over the past 10-15 years, to the study of non-traditional fields that are now designated as sub-disciplines: hospital, drug, radiation, genetic, molecular, nutritional, environmental, occupational, accidental injury, geriatric, and psychiatric, behavioral, and social epidemiology. Beyond these are the new frontiers of health policy, health services planning, health care financing, health care delivery, economic analysis, and health program evaluation that are increasingly relying on epidemiologic data and epidemiologic methodology. These will constitute the main growth area of the discipline in the 21st century. The paper identifies the main elements of the methodology that have remained constant and are likely to guide the growth of the discipline in the next century. The epidemiological methodology in essence is inference on the causal relation between cause and effect while preventing, eliminating, or adjusting for the 3 main types of bias (selection,misclassification, and confounding). Developments in the fields of statistical modeling with the accompanying hard and software will enable a deeper and richer understanding of multi-causal relations by describing interaction effects (synergistic and antagonistic) as well as effect modifications. Intervention, definable as looking at the causal relation in reverse, is the main underlying objective of epidemiological studies and involves interrupting, reversing or mitigating the effect of causes.Epidemiology will continue enjoying pride of place among public health disciplines in being able to devise ways of successful intervention even before the detailed understanding of the causal relation at the molecular level is achieved. The most effective interventions in this regard are those that modify human behavior and lifestyle. A preponderant majority of disease conditions can be either prevented or cured by available medical knowledge and technology. The last frontier that medicine has yet to conquer is changing the human will to adopt healthy life-styles and avoid unhealthy ones. Epidemiological studies can link specific life-styles to specific diseases enabling effective interventions even if the mechanisms may not be known fully. It is the considered opinion of the author that the increasing specialisation within epidemiology will not lead to the 'death' of the mother discipline. Neither will the mother discipline turn into theoretical epidemiology. It will remain close to its practical and empirical base. Epidemiologists will continue being involved in the application of their methodology to real-life public health problems unlike theoretical physicists or mathematicians who can afford to keep their hands clean from toiling in the field. Most of the growth forecast in epidemiology will be in the area of applying epidemiologic data to large and detailed data sets on human populations and environmental variables. No spectacular developments are forecast in the methodology of study design or study analysis. Meta or pooled analyses will become more popular as studies and databases proliferate.