Validity of Prediction of a Questionnaire of Knowledge and Attitudes towards Family Medicine

Abstract

Objective: to learn about the validity of prediction of a questionnaire that assesses knowledge and attitudes of students of medicine towards Family Medicine (fm). Material and methods: cohort study; the studied individuals were students enrolled in a Primary Care course (pc) between 2005 and 2006, who graduated in 2010 and chose a specialty in 2011. It was designed a questionnaire with 34 closed questions. Students were invited to answer before and after taking the pc course, as well as those students enrolled in the sixth course (2009-2010). The specialty chosen was registered; to calculate the predictive validity of the questionnaire, their sensitivity and specificity was estimated, analyzing the roc (receiver operating characteristic) curves. Results: the questionnaire was answered by 44 students before studying the course and 79 after studying it. At the end of the sixth course, 76 students answered the questionnaire. The area under the roc curve was, respectively, 0.299 (ic 95%: 0.094-0.504), 0.482 (ic 95%: 0.280-0.684) and 0.688 (ic 95%: 0.282-1.093). Only in the latter case the curve roc was above the diagonal. The best cut-off point was 31, with sensitivity=0.75 and specificity=0.714. Conclusions: the questionnaire has a poor validity of prediction concerning the choice of fm. It only improves during the last year of the degree.