Improving doctors' prescribing behaviour through reflection on guidelines and prescription feedback: a randomised controlled study.

Background: It is difficult to put research findings into clinical practice by either guidelines or prescription feedback.

Aim: To study the effect on the quality of prescribing by a combined intervention of providing individual feedback and deriving quality criteria using guideline recommendations in peer review groups.

Methods: 199 general practitioners in 32 groups were randomised to participate in peer review meetings related to either asthma or urinary tract infections. The dispensing by the participating doctors of antiasthmatic drugs and antibiotics during the year before the intervention period provided the basis for prescription feedback. The intervention feedback was designed to describe the treatment given in relation to recommendations in the national guidelines. In each group the doctors agreed on quality criteria for their own treatment of the corresponding diseases based on these recommendations. Comparison of their prescription feedback with their own quality criteria gave each doctor the proportion of acceptable and unacceptable treatments.

Main outcome measure: Difference in the prescribing behaviour between the year before and the year after the intervention.

Results: Before intervention the mean proportions of acceptably treated asthma patients in the asthma group and urinary tract infection (control) group were 28% and 27%, respectively. The mean proportion of acceptably treated patients in the asthma group was increased by 6% relative to the control group; this difference was statistically significant. The mean proportions of acceptable treatments of urinary tract infection before intervention in the urinary tract infection group and asthma (control) group were 12% for both groups which increased by 13% in the urinary tract infection group relative to the control group. Relative to the mean pre-intervention values this represented an improvement in treatment of 21% in the asthma group and 108% in the urinary tract infection group.

Conclusions: Deriving quality criteria of prescribing by discussing guideline recommendations gave the doctors a basis for judging their treatment of individual patients as acceptable or unacceptable. Presented with feedback on their own prescribing, they learned what they did right and wrong. This provided a foundation for improvement and the process thus instigated resulted in the doctors providing better quality patient care.