Written simulation of patient-doctor encounters. 2. Assessment of the performance of general practitioners

Two rating procedures were used to judge the performance of 19 general practitioners confronted with a written simulation of patient-doctor encounters. The simulation comprised five patients with vague complaints. A group of three 'expert' general practitioners judged the attention that the doctors paid to somatic aspects or to causes of the complaints. A second group of 18 experts judged the extent to which the therapeutic procedures of the general practitioners might induce a risk of unnecessary harm to the patient. Both rating procedures were shown to be reliable for three of the five simulated patients. A weak correlation between these issues was established for these three patients. The problems of judging the behaviour of the practitioners with the other two patients are discussed. The performance of the general practitioners was relatively constant. Variability between individuals substantially exceeded variability within an individual with respect to the attention given to somatic aspects, but these variabilities were roughly equal with respect to the risk of causing unnecessary harm.