In search of a solution to the primary care denominator problem.

Before morbidity or utilization rates can be legitimately compared across practices or within a practice over time, it is necessary to know the number and age-sex distribution of individuals who are served by the practice in question. Estimating this "population at risk" has been referred to as the "denominator problem." Although a variety of methods for estimating practice denominators were proposed more than six years ago, none has been shown to be satisfactory, and no new ones have materialized. One method, however, has never been evaluated, and evidence is presented which suggests that this method may be capable of providing satisfactory estimates of practice denominators. Data from the US Health Interview Survey, the British National Morbidity Study, and other sources suggest that it may be possible to derive age- and sex-specific correction factors which, when applied to the age-sex distribution of visiting patients, would provide reasonable estimates on the practice denominator. This "correction factor method" would require only that practices maintain age-sex registers of their visiting patients. Further investigation is required before this method can be considered a satisfactory means of estimating practice denominators.