Involving patients in primary care consultations: assessing preferences using discrete choice experiments

BACKGROUND: Shared decision making (SDM) involves patients and doctors contributing as partners to treatment decisions. It is not known whether or to what extent SDM contributes to the welfare arising from a consultation, and how important this contribution is relative to other attributes of a consultation. AIM: To identify patient preferences for SDM relative to other utility bearing attributes of a consultation. DESIGN OF STUDY: In parallel with a randomised trial in training GPs in SDM competencies and risk communication skills, a discrete choice experiment exercise was conducted to assess patients' utilities. SETTING: Twenty general practices in South Wales, UK. METHOD: Five hundred and eighty-four responders from 747 patients attending the randomised trial (response rate = 78%). All patients had one of four conditions (atrial fibrillation, menorrhagia, menopausal symptoms or prostatism) and attended a consultation with a doctor in their own practice. Patients were randomised to attend a consultation either with a doctor who had received no training in the study or risk communication training alone or SDM training alone, or both combined. RESULTS: Five key utility bearing attributes of a consultation were identified. All significantly influenced patient's choice of preferred consultation style (P<0.001). Larger increases in utility were associated with changes on "doctor listens" attribute, followed by easily understood information, a shared treatment decision, more information and longer consultation. Utilities were influenced by whether the doctor had received risk communication training alone or SDM training alone, or both combined, prior to the consultations. The randomised trial itself had identified that the communication processes of these consultations changed significantly, with greater patient involvement in decision making, after the training interventions. CONCLUSION: Shared treatment decisions were valued less than some other attributes of a consultation. However, patient utilities for such involvement appeared responsive to changes in experiences of consultations. This suggests that SDM may gain greater value among patients once they have experienced it.