Design of a method for the evaluation of clinical competence in primary care

• Published in: Atención primaria Vol. 26; no. 9; p. 590
• Main Authors: Barragán, N, Violan, C, Martín Cantera, C, Ferrer-Vidal Cortella, D, González Algas, J
• Format: Journal Article
• Language: Spanish
• Published: Spain 30.11.2000
• Subjects: Clinical Competence; Communication; Cross-Sectional Studies; Diagnosis, Differential; Female; Humans; Hypertension - diagnosis; Hypertension - therapy; Male; Medical History Taking; Patient Simulation; Physical Examination; Physicians, Family - standards; Primary Health Care
• ISSN: 0212-6567
• DOI: 10.1016/s0212-6567(00)78729-4

To measure clinical competence in primary care (PC). Descriptive cross-sectional study. Six PC teams in Barcelona. 25 family doctors volunteering out of a total of 60. Three simulated visits took place on three different days, allocated at random as a prior appointment for hypertension. Clinical competence was measured through the Miller criteria, a method with open short-reply questions combining the clinical interview with standardised patients (SP), review of clinical histories (CH) and doctor's self-evaluation (DS). The following components of competence (CC) were evaluated: anamnesis, physical examination, further tests, differential diagnosis, management, clinical history and communication. 22 doctors completed the study. Each CC was weighted by a group of experts, who also agreed two minimum levels of competence, 50% and 60%. Each doctor's final score was the result of the addition of each of the three cases divided by the maximum score possible. The CC values were always higher with SP than otherwise. 63.3% of doctors surpassed the 60% mark, and 90.9% the 50% mark. Internal consistency, measured with Cronbach's alpha, was 0.94, 0.73 and 0.67 for each clinical case. The kappa reproduction test was 0.25. The use of SP is a good method of evaluation, as it enables anamnesis, management of the patient, physical examination and communication to be assessed. The study is very reliable, with low reproduction, which is probably due to the few cases used.