Primary Care Consultations About Medically Unexplained Symptoms: Patient Presentations and Doctor Responses That Influence the Probability of Somatic Intervention

Abstract

In primary care, many consultations about physical symptoms that the doctor thinks are not explained by physical disease nevertheless lead to somatic interventions. Our objective was to test the predictions that somatic intervention becomes more likely a) when doctors provide simple reassurance rather than detailed symptom explanations and do not help patients discuss psychosocial problems and b) when patients try to engage doctors by extending their symptom presentation. Consultations of 420 patients presenting physical symptoms that the doctor considered unexplained by physical disease were audio-recorded, transcribed, and coded. Analysis modeled the probability of somatic intervention as a function of the quantity of specific types of speech by patients (symptomatic and psychosocial presentations) and doctors (normalization, physical explanations, psychosocial discussion). Somatic intervention was associated with the duration of consultation. Controlling for duration, it was, as predicted, associated positively with symptom presentations and inversely with patients' and doctors' psychosocial talk. The relationship with doctors' psychosocial talk was accounted for by patients' psychosocial talk. Contrary to predictions, doctors' normalization was inversely associated with somatic intervention and physical explanations had no effect. Somatic intervention did not result from the demands of patients. Instead, it became more likely as patients complained about their symptoms. Facilitating patients' psychosocial talk has the potential to divert consultations about medically unexplained symptoms from somatic interventions. To understand why such consultations often lead to somatic interventions, we must understand why patients progressively extend their symptom presentations and why doctors, in turn, apparently respond to this by providing somatic intervention.