You are viewing the site in preview mode

Skip to main content


Table 2 Audit and Feedback Interventions to decrease benzodiazepine use

From: Improving the use of benzodiazepines-Is it possible? A non-systematic review of interventions tried in the last 20 years

Study Location Intervention Design: Audit and Feedback Interventions targeting GPs Study Design and Size Result Follow up
Baker et al, 1997 [66] Leicester, UK Audit on all long-term users (> 4 weeks) in the medical centre then GPs received either: 1 = feedback on prescribing practices + criteria for the management of long term BZ users. 2 = feedback + criteria + reminder cards for patient files. RT 18 practices patients = 2409 long term BZ users Both groups changed after intervention with respect to levels of compliance to criteria. 8.2% of patients were stopped and 1.3% were decreasing BZs. No difference between groups. 2nd audit completed 1 year post intervention
Holden et al, 1994 [67] Liverpool, Southport - UK Audit of BZ use + GPs invited to 2 meetings on auditing BZ use in general practice. Individual practices determined their own BZ policy for prescribing and reducing use. Observational 15 practices, 3234 patients Overall reduction of 16%. Sig reduction in those <65 (25%) compared to those >65 = 12%. 2nd audit at 8 months (end of study)
Pimlott et al, 2003 [68] Canada - Ontario 1 = audit and feedback on GPs prescribing of BZs compared to peers and best practice + information sheet on BZs every 2 months for 6 months. 2 = Control group had the same intervention for antihypertensives. RCT 168 GPs (intervention) 206 GPs (control) No sig decrease in BZ prescribing and no sig difference between intervention and control groups. 6 months post intervention
Study Location Intervention Design: Audit and Feedback Targeting LTC Study Design and Size Result Follow up
McClaugherty, 1997 [69] Texas LTC pharmacist audited BZ use + gave feedback to nurses and doctors. Nurses were given sleep promoting guidelines. OT's & physio's were encouraged to increase activities for those who couldn't sleep. Quasi-Experimental 10 Nursing Homes, 3 Texas counties % of patients prescribed routine BZ decreased from 4.5% (baseline) to 1.6% (post intervention). % of patients prescribed BZ on an as needed basis increased from 7.9% (baseline) to 9.3% (post intervention) 3 months post intervention
Gill et al, 2001 [70] Ontario, Canada Review of patients chart + a letter was sent to the treating doctor if inappropriate e.g. long acting BZ explaining why medication was inappropriate and suggestions for alternative therapy. Quasi-Experimental 1 LTC facility, 450 Patients 37.9% of inappropriate medications were withdrawn or changed after the letter. 2 months after follow-up letters
Elliot et al, 2001 [37] Australia Audit and 1 h meeting = feedback to all staff on prescribing compared to other hospitals and review of literature + posters in wards Quasi-Experimental 9 hospitals (6 aged care 3 medical wards) No sig reduction in BZ use. Sig increase in appropriate prescribing at 8 week (22%) and 6 months (30%) post intervention. 4-8 weeks (all) and 6 months (for 3 hospitals only) post intervention
Study Location Intervention Design: Audit and Feedback + Education Targeting LTC Study Design and Size Result Follow up
Roberts et al, 2001 [34] QLD + NSW, Australia 1 = 11 hrs of problem based education session for nurses + wall charts, bulletins, telephone, visits. Written drug review for 500 selected patients. Report on review placed in patient's records and available to the GPs. 2 = Control RCT 52 nursing homes, 13 (intervention), 39 (control). Sig difference in the reduction of BZs between intervention (decreased 597 items/year/1000 residents) and control (increased 278 items/year/1000 residents). 12 months (end of study)
Batty et al, 2001 [35] England/ Wales Audit then: 1 = lecture to staff on literature review on appropriate prescribing of BZs. Feedback on prescribing compared to another hospitals. 2 = Bulletin (2 sided A4) with same information as lecture. 3 = Control RCT Elderly inpatients at 17 hospitals (6 lecture, 4 bulletin, 7 control) No sig change in any group but verbal group increased appropriate prescribing by 15%, bulletin decreased appropriate prescribing (9%) and control remained the same. 4-6 weeks post intervention
Eide and Schjott, 2001 [33] Norway 1 = Audit of BZ use, feedback to staff (reports and a presentation). Academic education to all staff by pharmacist, consisting of 6 simple rules for the use of hypnotics (data collected in 1995 and 2000). 2 = Control (data collected in 2000 only) CT 10 LTC Facilities, 5 (intervention) and 5 (control) Sig dif in the % of patients use BZS in control (44%) compared to intervention (24%) post intervention. Sig higher dose of BZs in intervention group in 2000 (60%) compared to 1995 (38%). 5 years post intervention
Crotty et al, 2004 [71] Adelaide, Australia Audit then: 1 = GP received education and guidelines and audit of use. Nurses received education in behaviour management and all staff received education on reducing psychotropic medication use 2 = control MRP 20 LTC facilites, 10 (intervention) and 10 (control) No sig reduction in BZ use (6.3%, intervention, 0% control), no significant decrease in long acting BZs (2.8% intervention and 0.9% control) and no sig difference in BZ being prescribed on a as needed basis (4% intervention and 1% control) 2nd audit was at 7 months (end of study)
  1. BZ = benzodiazepine, GP = General Practitioner, Av = Average, Sig = statistically significant (p < 0.05), LTC = long term care, OT's = Occupational Therapists, RCT = Randomized controlled trial, CT = Controlled trial, RT = randomized trial, MPR = Matched pair randomisation