Purpose: To explore the functioning of the agency relationship in primary care in the UK, using the clinical example of the decision to commence drug treatment to prevent coronary disease.

Methods: A well functioning agency relationship will see the physician providing information to the patient on alternative treatment options, in order that the patient can make an informed choice. This study recruited individuals likely to be at high coronary risk from 13 UK primary care practices. People were invited for risk screening and data on preferences to receive treatment were collected through two face-to-face interviews. At the first interview, participants were asked if they would choose treatment that reduced their coronary risk by 30%, assuming a number of different pre-treatment five-year coronary risks. Following the first interview, and on the basis of known clinical details of the patient, five-year coronary risk was calculated for each patient. At the second interview participants were told their risk and asked to indicate again whether they wished to receive treatment. Coronary risk and risk reduction were expressed in words, in figures and using visual aids, and comprehension was tested. Following the second interview, patients had a clinical consultation with their own general practitioner, specifically to consider their use of preventive drug treatment. Participating general practitioners were also interviewed.

Results: A total of 202 patients were recruited (177 male; median age 65 years; 70 taking long-term medication). All main ethnic groups and social classes were represented. We have complete data 197 patients. When informed of their actual coronary risk, 98 participants (50%) stated that they wished to commence therapy as a preventive measure. However, following the clinical consultation, drug treatment was commenced in only 18 patients. The qualitative interview data shed some light on the reasons for this discrepancy: concerns relating to the life-long nature of the medication, disparate views on the appropriate role of agency and shared decision-making, and reluctance by patients to take responsibility for treatment choices.

Conclusions: In the context of patients attending for coronary risk screening, this research suggests that the agency relationship is not working well. In this carefully constructed scenario, where patients were given detailed information on risks and treatments, it appears that patients' expressed views were largely over-ridden in the clinical consultation.