Validation of the patient roster in a primary care practice.

Knowledge of the size of a practice population is an essential base for the evaluation of new forms of health care delivery and for epidemiologic research in primary care. Remuneration to providers in Ontario's Health Service Organization and Health Centre programs is partially based on the number of people listed on the patient roster as members of the practice. However, the accuracy of these rosters has never been determined. A mail and telephone survey was conducted to validate the roster in one such health center. A random sample of 1,065 households was contacted and a 78 percent response rate was obtained. The practice roster showed a population of 3,134. The age- and sex-adjusted estimate from the survey was 2,964 (+/- 262) individuals. The sensitivity, specificity, and accuracy of the roster were 0.90 or greater. It is concluded that this particular roster is a valid indicator of practice size, but caution is expressed about generalizing these results to other practices.