The development and evaluation of a computerised decision support system for primary care based upon 'patient profile decision analysis'.

Objective: To develop and evaluate in primary care a computerised decision support system for the management of stroke patients based upon 'patient profile decision analysis'.

Design: The decision support system incorporated the findings of 960 Markov models examining the decision to prescribe aspirin in the secondary prevention of stroke. The models reflected each combination of nine risk factors that determined a patient's profile. The evaluation comprised a qualitative interview and a questionnaire administered before and after the general practitioners (GPs) were given access to the support system.

Setting: Primary care.

Participants: 15 GPs from the West Midlands.

Main Outcome Measures: Decision certainty scoring of hypothetical patient vignettes. Qualitative perceptions of the applicability and acceptability of the system for primary care.

Results: After using the system, GPs were more certain of their decision making and made decisions more in line with national guidelines. Quantitative results further suggested that the system made decision making easier, improved feelings of being supported, improved the quality of decision making and increased satisfaction. Qualitative themes included that GPs thought the system could clarify their own decision making and improve GP-patient dialogue.

Conclusions: The feasibility of individualised decision analysis for general practice has been questioned. Patient profile decision analysis, however, may be a valuable means of harnessing some of the advantages of the methodology to produce more patient-specific guidelines for primary care.