From Wiki Journal Club
Jump to: navigation, search
Tannenbaum C, et al. "Reduction of inappropriate benzodiazepine prescriptions among older adults through direct patient education: the EMPOWER cluster randomized trial". JAMA Internal Medicine. 2014. 174(6):890-898.
PubMedFull text

Clinical Question

In elderly patients on chronic benzodiazepine therapy, can a direct patient education program decrease benzodiazepine usage?

Bottom Line

In elderly patients, an educational program resulted resulted in increased benzodiazepine discontinuation and decreased daily dosage.

Major Points

Benzodiazepines are potentially inappropriate medications in the elderly and several campaigns have be aimed at prescribers and patients regarding their potential for cognitive impairment, falls risk, and dependence. The Eliminating Medications Through Patient Ownership of End Results (EMPOWER) trial enrolled 303 long-term benzodiazepine users from 30 community pharmacies and randomized 148 patients from 15 pharmacies to educational intervention and 155 patients from 15 community pharmacies to a "wait list" control. The education intervention consisted of an 8 pages booklet regarding the harms of benzodiazepine usage and a 21 day tapering protocol.

At 6 months, there was a significant reduction in benzodiazepine usage (27% vs 4.5% discontinuation rate) and dosage reduction (11% vs 6%).


  • Cluster randomized triple blinded controlled trial
  • n = 303 patients, 30 community pharmacy clusters
    • educational intervention: n = 146, 15 pharmacies
    • wait-list control: n = 155, 15 pharmacies
  • setting: 30 community pharmacies in the Greater Montreal Area, Canada
  • enrollment: June 2010 to May 2013
  • follow-up: 6 months
  • analysis: intention to treat
  • primary outcome: % discontinuation of benzodiazepine therapy
  • secondary outcome: % discontinuation or dosage reduction of benzodiazepine


Inclusion Criteria

  • community pharmacy:
    • 20% of patient population consistent of adults 65 or older
    • ≥ 50 eligible participants
    • part of retail chain with 200 km of research center
  • patients:
    • ≥ 65 year old adults living in the community
    • ≥ 5 medications
    • ≥ 3 months use of potentially inappropriate medications

Exclusion Criteria

  • severe mental illness, dementia, epilepsy; concomitant usage of
    • antipsychotic medication (e.g. lithium, olanzapine, quetiapine, haloperidol, etc)
    • anticonvulsant medication (e.g. carbamazepine, valproate, gabapentin, etc)
    • cholinesterase inhibitor or memantine

Baseline Characteristics

From the intervention group:

  • mean age: 75 (SD 6.5, range 65 - 91)
  • female: 70.3%
  • college/university education: 21.6%
  • lives alone: 46.6%
  • self-reported fair or poor health: 35.8%
  • Montreal Cognitive Assessment Score: 25.4 (SD 24. range 21 - 30)
  • self-reported indication for benzodiazepine use
    • insomnia: 60.8%
    • anxiety: 45.9%
    • pain: 2.7%
    • other: 6.8%
  • anxiety disorder (Geriatric Anxiety Index ≥ 9): 32.4%
  • daily benzodiazepine dosage (lorazepam equivalent) 1.2 mg (SD 0.8, range 0 - 4.8)
  • benzodiazepine type
    • short acting: 29.1%
    • intermediate acting: 66.2%
    • long acting: 4.7%
  • duration of benzodiazepine usage: 9.6 years (SD 9.7, range 0.3 - 48)
  • previous cessation attempt: 45.2%
  • number of daily medications: 9.9 (SD 3.9, range 4 - 24)


Those randomized to the intervention group were provided with an 8 page booklet (6th grade reading level, 14 point font) containing:

  • self-assessment regarding the risk of benzodiazepine use
  • evidence for benzodiazepine induced harms and drug interactions
  • knowledge statements designed to induce cognitive dissonce about the safety of benzodiazepine use
  • peer champion stories designed to improve self-efficacy
  • 21 day tapering protocol and recommendations for insomnia/anxiety therapeutic substitutes
  • recommendation to discussing deprescribing regimens with the patient's physician or pharmacist
  • customized with the patient's name and benzodiazepine

The intervention was mailed to participants within 1 week of allocation, while the usual care group received the tool 6 months following allocation.


Presented as intervention vs usual care.

Complete cessation was defined as absence of benzodiazepine renewal at 6 month that was sustained for ≥ 3 months without a switch to another benzodiazepine.

Dose reduction was defined as ≥ 25% dosage reduction compared to the 6 month baseline before randomization that was sustained for ≥ 3 months.

The odds ratio was adjusted for age, sex, education, health status, number of medications, indication of benzodiazepine use for insomnia, anxiety disorder, benzodiazepine dose, duration of use, and previous attempt attempt at tapering.

Primary outcome

Complete cessation of benzodiazepine

27% vs. 4.5%, (adjusted OR 8.33, 95% CI: 3.32 - 20.93, NNT = 4.35)

Secondary outcomes

Cessation or reduction in benzodiazepine usage

37.8% vs 11% (adjusted OR 5.49, 95% CI 2.78 - 10.84, NNT 3.70)

Other outcomes

Participants in the intervention group who completed the trial (n =123) were interviewed via telephone.

initiated discussion with physician or pharmacist: 62%

attempted discontinuation: 58%

  • successful: 54%
    • received substitute therapy: 13% (trazodone, amitriptyline, paroxetine reported)
  • achieved dose reduction: 22%
    • 1/3 still tapering at study end
  • failed: 24%
  • experienced withdrawal symptoms: 42%
  • followed tapering protocol provided: 72%

reasons for not wanting to attempt to taper

  • discouraged by physician or pharmacist 33%
  • fear of withdrawal symptoms: 25%
  • lack of concern over taking benzodiazepines: 23%
  • difficult life circumstances: 12%


Funded by a grant from the Canadian Institutes of Health Research.

Sponsored by Centre de Recherche de l'Institut Universitaire de Geriatrie de Montreal

Authors retained full independence from the study sponsor in terms of

  • study design and conduct
  • data collection, management, analysis and interpretation
  • preparation, review, and manuscription approval
  • decision on whether to submit for publication


  • limited timeframe for followup (potential for relapse or for patients to still be tapering at study end)
  • excluded patients with dementia
  • low recruitment rates for pharmacies (18%) and individual participants (11%)
  • study did not evaluate substitution of benzodiazepines with other potentially inappropriate therapies, such as Z-drugs or atypical antipsychotics

Further Reading

American Geriatric Society 2015 Update Beers Critieria for Potentially Inappropriate Medication Use in Older Adults. J Am Geriatr Soc. 2015;63(11);2227-46)

Canadian Guideline for Safe and Effective Use of Opioids for Chronic Non-Cancer Pain. Appendix B-6: Benzodiazepine Tapering. Canada: National Opioid Use Guideline Group (NOUGG); 2010