A systematic review of pharmacist-led audit and feedback interventions to influence prescribing behaviour in general practice settings

Abstract

Introduction: Audit and Feedback (A&F) involvesmeasuring data about practice, comparing it with clinicalguidelines, professional standards or peer performance, andthen feeding back the data to individuals/groups of healthprofessionals to encourage change in practice (if required).A 2012 Cochrane review (1) found A&F was effective inchanging health professionals' behaviour and suggested thatthe person who delivers the A&F intervention influences itseffect. Increasingly, pharmacists work in general practice andoften have responsibility for medication review and repeatprescriptions. The effectiveness of pharmacist-led A&F ininfluencing prescribing behaviour is uncertain.Aim: This secondary analysis from an ongoing update ofthe original Cochrane review aims to identify and describepharmacist-led A&F interventions and evaluate their impacton prescribing behaviour in general practice compared withno intervention.Methods: This sub-review is registered with PROSPERO:CRD42020194355 and complies with the PreferredReporting Items for Systematic Reviews and Meta-Analyses(PRISMA) guidelines (2). For the updated Cochrane review,the Cochrane Effective Practice and Organization of CareGroup searched MEDLINE (1946 to present), EMBASE,CINAHL and Cochrane Library (March 2019) to identifyrandomised trials featuring A&F interventions. For this subreview, authors screened titles and abstracts (May 2020) toidentify trials involving pharmacist-led A&F interventions inprimary care, extracted data, and assessed risk of bias (RoB)in eligible studies. Review results are summarised descriptively. Heterogeneity will be assessed and a random-effectsmeta-analysis is planned. Publication bias for selected outcomes and the certainty of the body of evidence will be evaluated and presented. Sub-group analyses will be conducted.Results: Titles and abstracts of 295 studies identified forinclusion in the Cochrane A&F review update were screened.Eleven studies (all cluster-randomised trials) conducted in 9countries (Denmark, Italy, Netherlands, Norway, Republicof Ireland, UK, Australia, Malaysia, USA) were identifiedfor inclusion (Figure 1). Six studies had low RoB, two hadhigh risk due to dissimilarities between trial arms at baseline and/or insufficient detail about randomisation, andthree studies had unclear RoB. Studies examined the effectof A&F on prescribing for specific conditions (e.g. hypertension), medications (e.g. antibiotics), populations (e.g.patients >70), and prescribing errors (e.g. inappropriatedose). The pharmacist delivering A&F was a colleague ofintervention participants in five studies. Pharmacists' levelsof skill and experience varied; seven studies reported details of pharmacist training undertaken for trial purposes.A&F interventions in nine studies demonstrated changes inprescribing, including reductions in errors or inappropriateprescribing according to the study aims and smaller increases in unwanted prescribing compared with the controlgroup. Data analyses are ongoing (results will be availablefor the conference).Conclusion: The preliminary results demonstrate the effectiveness of pharmacist-led A&F interventions in differentcountries and health systems with influencing prescribingpractice to align more closely with guidance. Studies measured different prescribing behaviours; meta-analysis is unlikely to include all 11 studies. Further detailed analysisincluding feedback format/content/frequency and pharmacist skill level/experience, work-base (external/internal torecipients), will examine the impact of specific features onintervention effectiveness.