Purpose: To compare changes in self-assessed clinical confidence over a two-year residency between two groups of family practice residents, one starting in a family practice center and the other starting in a hospital.
Method: All 44 entering residents at the University of Western Ontario in either 1984 or 1985 were eligible. Forty-two participated at baseline, and 24 (57%) provided completed responses after two years. Confidence regarding 177 topics in 19 general topic areas was assessed using self-completed questionnaires administered at baseline and after six, 12, and 24 months. The residents rotated every six months between sites, with approximately half starting in each site. Nonrandom assignment to starting site included consideration of the residents' stated preferences; hence self-selection bias was possible. Mean differences were evaluated using t-tests, and trends over time were assessed using repeated-measures analysis of variance.
Results: There was no difference between the groups' grand means at any time. In the 19 areas, five statistically significant differences were seen, four at six months and one at 12 months; four involved greater confidence by the center-start group. No difference was seen after two years, and both groups displayed substantial increases in confidence over two years.
Conclusions: The few significant differences in confidence between the two groups suggest that their training might have been deficient if it had omitted one of the settings. However, because these differences had disappeared after the groups had each experienced 12 months in each setting, the initial site of training appears not to affect learning.