APPROACHES TO THE DENOMINATOR IN PRACTICE-BASED EPIDEMIOLOGY - A CRITICAL OVERVIEW

Objectives-An accurate knowledge of the population at risk is a fundam ental requirement for determining rates and making comparisons in epid emiological research. The major obstacle of studying the epidemiology of sentinel practice networks is the determination of population at ri sk, in this case, the reference population of medical practices. This article is intended to give a brief overview of major denominator appr oaches used in practice based epidemiology today, to discuss their und erlying assumptions, their strengths and Limitations. Design-The liter ature used in this paper was searched from Medline databases of 1970-1 997 using the logical expression ''denominator and practice''. More li terature was identified from the references cited in those articles an d from research reports that were available to the authors. Main resul ts-There are various approaches to the denominator at different levels of complexity, which are presented akin to the well known ''iceberg p henomenon'': with only a small portion of the iceberg visible above th e surface, inference as to the size of the invisible part may still be made under certain assumptions. Crude numbers of cases may still refl ect trends in the true epidemiology of disease and may be useful for t ime-series analyses. Differences in the number of network participants over time and across region may be controlled for by using the number of sentinel practices as a denominator. The number of consultations i s a first step towards a population-based denominator, reflecting char acteristics of both patients and the network. The yearly or quarterly contact group is a true person-based denominator, yet disregarding the population not consulting. The population in practices' catchment are as can be either determined from patient lists or estimated using math ematical models. The ideal denominator is the total population in a ge ographically defined area, though this information can be directly rel ated to medical practices only in very few countries. Conclusions-Alth ough a person, or ideally a population-based denominator is desirable, even ''lower-level'' denominators may be suitable for certain researc h topics. In countries without patient registration, the estimation of incidences and prevalences has many methodological uncertainties that limit the use of sentinel practice systems. Assuming representativene ss, valid analytical or time-series studies, however, can still be car ried out even if there is very little information on the population at risk covered by particular medical practices.