A randomised trial of prescribing feedback to general practitioners using academic detailing versus postal bulletin

C. Naughton, J Feely, K. Bennett. Department Pharmacology & Therapeutics, Trinity College, Dublin Ireland

Background:Academic detailing (15-20 min face to face interviews discussing individual GP prescribing) is often used as part of a multifaceted intervention to promote rational prescribing (O’Brien et al 1997). However academic detailing incurs greater costs and its additional effectiveness needs to be evaluated compared to conventional methods such as audit and feedback.

Objectives: To compare the effectiveness of a randomised intervention using academic detailing versus postal bulletin to:
1) Reduce antibiotic prescribing rates
2) Increase selection of first line antibiotics (penicillin) over second line antibiotics (co-amoxiclav and cephalosporin)
3) Increase generic prescribing for five targeted drugs
4) Increase prescribing of secondary preventative therapies in patients with cardiovascular disease (CVD) or diabetes

Methods:Over one third (110/300) of eligible GPs from the Eastern Region of Ireland and involved in the HSE-Primary Care Reimbursement Service (HSE-PCRS) took part in the study. These GPs were randomised to either group 1 (n=55) receiving individualised prescribing feedback as a postal bulletin or group 2 (n=55) each GP received the postal bulletin plus an academic detailing visit. The visits lasted 15-20 minutes in which the individualised prescribing feedback was discussed. Each GP received two separate bulletins +/- two visits. GP prescribing and demographic data were obtained from the HSE-PCRS database. Log linear regression was used to compare six month antibiotic and CVD prescribing rates between the groups adjusting for baseline prescribing and patient population structure. Segmented regression analysis of interrupted time series was used to examine 12 month pre and post intervention changes in generic prescribing trend between the groups. P values are presented and significance at p<0.05 is assumed.

Results:There were no significant differences between the randomised groups in term of GP characteristics or patient panel structure.There was no significant reduction (p=0.85) in total antibiotic prescribing at six months post intervention in the academic detailing group compared to bulletin only GPs, and academic detailing GPs were no more likely to prescribing first line antibiotics (p=0.34) than bulletin only GPs.

There was no significant difference between the groups in generic prescribing trends for the five targeted drugs in the 12 months post intervention period. e.g. pravastatin (p=0.31) or omeprazole (p=0.25). Similarly prescribing of secondary preventative therapy in patients with CVD (statins p=0.80, anticoagulants p=0.83) or diabetes (statins p=0.86, anticoagulants p=0.16) was not significantly different between the randomised groups.

Conclusion: Prescribing feedback using academic detailing was not significantly different than postal bulletin alone in changing GP prescribing behaviour. The non-significant result for academic detailing may indicate that this aspect of the intervention needs to be more intensive to achieve a significantly greater impact than prescribing feedback alone.

0’Brien MA, et al (1997) Cochrane database of systematic reviews CD000409.