Physician concurrence with primary care guidelines for persons with HIV disease

An anonymous questionnaire was mailed to 3 groups of physicians in British Columbia: 659 registrants of a province-wide human immunodeficiency virus (HIV) drug treatment programme (Group 1); 765 who had had a patient test positive for HIV (Group 2); and a random sample of 484 (Group 3). Questionnaires provided data about: physician's demographic and personal characteristics; HIV-related patient care experience; and 12 knowledge-based items. Responses were received from 65% of Group 1, 68% of Group 2 and 70% of Group 3, with complete information received from 38% of Group 1 and 50% of Groups 2 and 3. Summary knowledge scores were computed by comparing physician responses to recommendations of contemporary guidelines. Regression techniques identified associations between physician characteristics and knowledge scores. Multivariate analysis revealed an inverse relationship between knowledge and physician age in all groups (all P < 0.01). Increased knowledge was associated with the number of currently active HIV-positive patients in Groups 1 and 2 (all P < 0.001), and lack of specialization in Groups 2 and 3 (both P < 0.001).