Pap smears in general practice: a secondary analysis of the Australian Morbidity and Treatment Survey 1990 to 1991.

We investigated the characteristics of Australian general practice that predict performance of Pap smears by secondary analysis of the Australian Morbidity and Treatment Survey 1990 to 1991. Chi-squared analysis identified potential associations between Pap smear rate and patient, doctor and practice variables. Significant associations were examined using logistic regression and generalised estimating equations. Participants were 495 general practitioners who collected information on 113,468 doctor-patient encounters, of which 43,211 encounters involved females aged 18 to 70 years. Pap smear encounter (2449) were identified and classified as patient-requested (62 per cent), diagnostic (5 per cent) or opportunistic (33 per cent). The large difference in the unadjusted Pap smear rates per 100 female encounters for female general practitioners (11.7) and male general practitioners (4.2) required separate analysis by sex of the general practitioner. For male general practitioners, a Pap smear was less likely: as patient age increased; for new patients; for general practitioners with less general practice experience; for general practitioners with no postgraduate qualifications; with metropolitan practice location; and if the practice had more than 25 per cent of patients with English as a second language. For female general practitioners, a Pap smear was less likely: for older known patients; as the age of the general practitioners increased; and for management of fewer problems per 100 encounters. A Pap smear was less likely to be opportunistic: as patient age increased; for general practitioners who were Australian graduates; and for general practitioners with no postgraduate qualifications. Consideration of patient, doctor, and general practice characteristics may facilitate the design of interventions to improve cervical cancer screening.