Skip to main content

Table 1 Description of characteristics of included studies, significant results and related TDF domains

From: Barriers and enablers for deprescribing benzodiazepine receptor agonists in older adults: a systematic review of qualitative and quantitative studies using the theoretical domains framework

Author, year Country Type of study Population Measures/themes explored BZRA deprescribing definition Main extracted results
Studies conducted in the NH setting
Anthierens, 2009 [36] Belgium Qualitative 33 NH nurses, in 5 NHs Focus group and face-to-face interviews exploring BZRA use in NH, perceived role of nurses and their attitudes and feelings Not reported Different themes and subthemes:
Nurses individual attitudes and perceptions:
-Routine approach
-Nurse as ‘sleep guardian’
Pharmacological knowledge
Organisational factors:
-Traditional doctor-nurse relationship
-Organisational requirements
-Smooth running of the NH
Team meetings
Bourgeois, 2014
Study A [37]
Belgium Quantitative, BZRA deprescribing feasibility study,
135 NHRs in 5 NHs GPs and NHRs willingness to deprescribe
Feasibility of deprescribing
BZRA discontinuation Willingness for discontinuation:
GPs refused deprescribing for 71 NHRs
- Unmotivated patient: 40
- Previous attempt failed: 13
- Other reasons (too old): 9
- Medical cause of sleeplessness: 8
- BZRA not indicated for sleep problems: 6
- Psychiatric history: 5
- Discontinuation started before inclusion possibility: 4
- Stopped BZD but switched to something else: 2
- Discontinuation but no agreement to start in study: 2
13 NHRs out of 51 refused
Main reason: reluctance towards change
Feasibility of discontinuation:
Of 38 NHRs, 66.0% were successful
in completely discontinuing BZRA use at 8 months
The reasons why NHRs were not successful in their discontinuation was because of sleep problems during the discontinuation (n=6/13) and overall loss of motivation (n= 5/13)
Bourgeois, 2014
Study B [38]
Belgium Quantitative cross-sectional study: GP and nurse questionnaire 25 GPs and 16 nurses in 5 NHs Initiation, indication and previous attempts to stop chronic BZRA use
Benefit and harm of BZRA use
Willingness to stop BZRA use
Resident-specific barriers
General attitudes towards BZRA discontinuation
BZRA discontinuation Previous attempts
The GPs and the nurses indicated that they had already attempted a withdrawal in the past in 26% and 12% of the residents, respectively
In general, chronic BZRA use was perceived as effective, and only few adverse effects were noted
Willingness to stop
GPs and nurses indicated willingness to stop chronic BZRA use in respectively 33 and 21% of NHRs
Resident-specific barriers
most common barriers among the GPs were:
-Fear of resistance from the resident
-Preference of a pharmacological treatment above a non-pharmacological treatment
-Fear that in these resident initial problems would come back
-Fear of an increase in the care burden for the staff
-Perception that change is not necessary as long as the resident functions well
-Fear of withdrawal effects.
Among the nurses, the most common barriers were:
-Fear that the residents’ initial problems would come back
-Preference for a pharmacological treatment
-Conviction that change is not necessary as long as the resident functions well.
General attitudes
The most common attitudes among both GPs and nurses were:
-The longer the resident takes the medication, the more difficult it is to stop.
-Old age of a resident makes it difficult and unnecessary to stop.
-Help of other care professions, such as a psychologist and a pharmacist, are not really necessary.
-Alternative strategies are more time-consuming
-Rhythm of a nursing home with strict bedtimes also limits possibilities for discontinuation.
Nurses agreed and GPs disagreed on the statements that there is little knowledge on alternative strategies to cope with problems when stopping BZDs and that there is little scientific information available for stopping.
de Souto Barreto, 2015 [13] France Quantitative non-randomised controlled trial of implementation of an intervention (not targeted at BZRA) 3973 NHRs from 163 NHs,
2167 included in this substudy
Factors associated with BZRA deprescribing at 18 months BZRA discontinuation In general, no association for NH and NHR factors BUT living in a particular NH affected BZRA deprescribing
In the intervention group, use of an antidepressant was a facilitator and female gender was a barrier
In the control group, use of meprobamate and a higher number of medications were both barriers
Evrard, 2020 [39] Belgium Quantitative cRCT of an intervention (not targeted at BZRA) implementation 418 NHRs with BZRA at baseline, in 54 NHs Factors associated with BZRA deprescribing at 15 months Deprescribing included complete cessation or decreased daily dose and cessation of an “if needed” BZRA prescription in addition to an unchanged chronic dose Enablers:
-Intervention (consisting of education programme for HCPHCPPs, multidisciplinary work and medication review)
-Hospitalisation in the past 3 months
-Parkinson/extrapyramidal syndrome
-Number of beds in the NH
-Public NH (as compared to private non-profit)
Lambson, 2003 [40] South Africa Mixed-methods (qualitative interview and quantitative questionnaire) Interviews: 44 older adults of a retirement village
10 GPs
4 nursing staff
Interview: reasons for initiation and continuation of BZRA, duration and frequency of use, perception of their doctor’s attitude and prescribing behaviour, desire and/or efforts to stop taking the tablets, and perceptions of usefulness
Questionnaire: BZRA perceptions and concerns
Not reported Patient interviews:
-Reasons to continue taking the medication related to fear of not sleeping without them
-47% of subjects felt that their doctors were quite happy for them to continue taking the benzodiazepine
-26% reported that they would like to stop taking BZRA
GP questionnaire:
-66% believed that it was acceptable to allow an elderly patient to continue a benzodiazepine indefinitely, and 78% felt that a regular sleeping tablet was a good idea.
-44.4% did not mind renewing prescriptions for benzodiazepines, while 33.3% did mind.
-55.5% felt it was easier to renew the prescription than to argue with the patient.
-79% agreed that patients taking sleeping tablets would not be persuaded to give them up.
-67% felt they were bothered by endless requests for benzodiazepine prescriptions
Nursing staff questionnaire:
-75% felt it was not a problem to continue the sleeping tablet indefinitely
-All subjects believed that it was better to tranquillise a restless patient than to allow them to disturb other patients.
MacLagen, 2020 [41] Canada Quantitative, retrospective cohort study 35 169 NHRs from Ontario Factors associated with BZRA discontinuation (in general and stratified by sex) ≥ 30-day gap in days supplied during the 180-day follow-up period after admission In general
-Female gender
-Low income
-Aggregated diagnosis group (more comorbidities)
-Older age
-Higher aggressive behaviour scale score
Among men
-Older age
-Higher aggressive behaviour scale score
Among women
-Low income
-Aggregated diagnosis group (more comorbidities)
-Older age
-Higher aggressive behaviour scale score
Mestres Gonzalvo, 2018 [42] Netherlands Quantitative, BZRA deprescribing clinical alert feasibility study, uncontrolled GPs received 180 alerts about 161 NHRs, in 15 NHs Reasons for not following a clinical alert related to BZRA deprescribing (as indicated by GPs)
Differences between followers and non-followers
Following the clinical rule meant that the GP started BZRA deprescribing 27 out of 161 (16.8%) of the alerts actioned were followed:
Reasons for not following:
-Already tried (n=10, 6.2%)
-Patient/family resistance (n=37, 23.0%)
-No need (non-continuous BZRA use) (n=32, 19.9%)
-Indication still present (n=27, 16.8%)
The alert-following group had a shorter median-time of BZRA, compared to the alert non-following group
Rate of alert following differed among GPs
Studies conducted in the ambulatory setting
Allary, 2020 [43] Canada Quantitative, RCT of gradual withdrawal programme with or without cognitive-behavioural therapy 60 patients of the PASSE-60 study, Predictors of successful BZRA deprescribing at the patient level in a 16-week RCT, at end of withdrawal, 3 months and 12 months later Complete BZRA discontinuation At end of withdrawal:
2 enablers:
-Higher social support satisfaction
-Lower BZRA dose at baseline
At 12 months:
4 enablers
-Higher support satisfaction
-Self-perceived competence
-Higher intensity of depressive symptoms
-Poorer quality of sleep
Barter, 1996 [44] United Kingdom Qualitative 11 older adults, chronic BZRA users Interviews about: type of benzodiazepine used; length of use and pattern of taking; social support; reasons for first using these tablets and current reasons; perception of doctor’s attitude and prescribing behaviour; wishes and efforts to stop taking the tablets; and general sleep quality Not reported Major themes:
-Reported efficacy of the prescribing sleeping tablets
-Changes in dosage/pattern of use (for current benzodiazepines)
-Reasons for use
-Patient knowledge and perception of doctor’s attitude and prescribing behaviour in relation to benzodiazepine use
-Wishes, efforts and past experiences of discontinuing tablets
-How people felt about using sleeping tablets
Bell, 2011 [45] Finland Quantitative, longitudinal (2 years) observational study 311 patients using BZRA Factors associated with BZRA deprescribing No benzodiazepine prescription in the last 6 months One enabler:
-Age over 85 years
Chen, 2010 [46] Canada Qualitative 13 HCPs working in a geriatric day hospital
5 patients who had been referred to the pharmacist for BZRA deprescribing
Providers: Interviews and group discussion of perceived and actual role in BZRA deprescribing, barriers and enablers, factors that predispose, enable and reinforce
Patients: Barriers and facilitators, factors that predispose, enable and reinforce, role of various providers
Not reported Providers:
-Physicians, nurses and pharmacists were the most involved
-Other providers lack of guidance
Patients, main themes:
-Experiences with BZRA
-Willingness to taper
-Tolerance of tapering
Chen, 2014
USA Quantitative, quasi experimental comparative study.
Intervention=loosing BZRA reimbursement
Intervention: 250 older adults
216 older adults
Factors associated with BZRA cessation after losing coverage, two regression models BZRA discontinuation (2 regression models used) In one of the two models used:
-Older age
-Higher comorbidity scores
-Higher BZRA exposure
-Higher BZRA dosage
-Gender: woman
Cook, 2007
Study A [48]
USA Qualitative 50 older adults, chronic BZRA user Interviews about: the rationale and circumstances for BZRA use; patient’s perceptions of family members and physicians’ perspectives; knowledge of adverse effects; experience of skipped doses; reliance on benzodiazepines; thoughts about discontinuation; interest in finding alternatives Not reported Different themes and subthemes:
Purpose and importance of benzodiazepines
-Means to cope with stress/anxiety and aid sleep
-Lifeline or life-transforming properties
-Lack of awareness, underestimation or disregard for adverse effects
Attitudes towards tapering/discontinuation
-Negativity or resistance to tapering/discontinuation
-Rejection of psychological interventions
Power and influence of physician-patient relationship
Cook, 2007 Study B [49] USA Quantitative,
46 older adults taking BZRA for at least 3 months Factors associated with willingness to taper/discontinue BZRA Tapering or discontinuation 2 barriers identified in multivariate analysis:
-More frequent benzodiazepine intake
-Higher Anxiety Sensitivity Index
Cook, 2007
Study C [50]
USA Qualitative 33 GPs Interviews about:
-Role of BZRA and management of anxiety, insomnia, and depression in older adults
-BZRA prescribing and renewing process
-Problems with BZRA use and strategies, including psychotherapy use
Not reported Different themes and subthemes:
Physician minimisation of the problem:
-No addiction seen in this population
-Little recognition of adverse effects other than addiction
-Continuation is compassionate, discontinuation is harsh
-Low priority relative to medical problems
Justification of short-term and long-term BZRA use:
-Effectiveness for anxiety and sleep problems
-Belief that stable dosage is safe and effective
-Attempt to discontinue will fail
-Anticipated resistance from patient
-Cost-benefit: Question patient gain and highlight suffering involved in tapering or discontinuation
Broad organizational factors and system constraints:
-Limited physician time
-Poor reimbursement for mental health care
-Older patients limited acceptance and access to mental health services
Iliffe, 2004 [51] United Kingdom Qualitative 192 patients, long-term (>6 months) BZRA users Interviews about:
-BZRA prescribing patterns
-Reasons for initial and current prescription
-Belief in BZRA efficacy
-Concerns about BZRA
- Previous discontinuation attempts
- Perceived advantages and disadvantages of deprescribing
Evaluate differences between continuers and withdrawers
Withdrawers= patients who wished to participate in a withdrawal programme Reason for prescription: No difference between continuers and withdrawers in reason for initial or current prescription
Beliefs of efficacy:
-Continuers reported BZRA as more helpful than withdrawers
-More continuers reported that they never had sleeping problems while taking BZRA, while withdrawers reported having problems very often
Concerns about BZRA:
-In both groups the majority of patients reported that no one ever suggested their BZRA may be harmful
-Only few patients had worried about long-term use adverse effects, but more withdrawers than continuers said they had worried
Perceived advantages and disadvantages of BZRA deprescribing:
-No difference between continuers and withdrawers regarding disadvantages (mainly not sleeping and not being able to manage)
-Withdrawers were more likely to mention specific benefits, including ‘clearer thinking’, ‘better memory’, ‘being more in control’, ‘more natural sleep’, ‘having to take less tablets’, ‘feeling less sleepy’, ‘feeling proud of myself’
Intention: Withdrawers were more willing than continuers to stop BZRA.
    83 practice staff: 72 GPs, 5 managers, 4 nurses, 2 counsellors Group interview on main problems, advantages and disadvantages they anticipated in withdrawing their elderly patients from BZRA   Prevalence of beliefs about BZRA deprescribing advantages and disadvantages:
-Increased demand on the GP: 48%
-Anticipated difficulty in persuading older adults: 51%
-Problems anticipated:
°Upset or anxious patients: 34%
°Patient insomnia: 24%
°Unmasked depression: 21%
-Expected benefits:
°Fewer falls: 57%
°Better sleep: 25%
°Better quality of life: 18%
°Increased independence and unmasking of depression: 11%
-Better clinical practice: 65%
-Reduced prescription costs: 34%
-Addressing a “significant public health issue”: 22%
In addition, respondents were concerned about “how patient would react to being encouraged to withdraw from ‘harmful’ drugs which had been prescribed by their own GP”
Joester, 2010 [52] Australia Quantitative, retrospective, cross-sectional study 42 BZRA users, over 65, attending a fall clinic Factors influencing recommendations for BZRA deprescribing
Factors associated with compliance
Successful dose reduction or discontinuation Recommendation to deprescribe BZRA in 31/42 patients
Enablers of BZRA deprescribing recommendation:
-Assessment by a geriatrician (compared to a rehabilitation specialist)
-Patients using BZRA as needed or less than three times per week
Compliance with recommendation in 21/28 patients
Enabler of recommendation compliance:
-Advise to cease BZRA completely (compared with advice to reduce dose or gradually reduce dose with the aim of cessation)
Kuntz, 2018 [53] USA Qualitative study, parallel to a direct-to-patient educative intervention 10 older adults, Z-drug users Interview about:
- Past and current Z-drug use
- Prior education about sedative
- Educational needs
- Reactions to the intervention material
Not reported 3 major themes:
-Insomnia-related factors (Importance of sleep, treatments alternatives)
-Structural and health care delivery system related factor (identification to brochure, individual care, no discussion about it with GP)
-Patient experiences and concerns (BZRA taking experience, side-effects)
    7 GPs Interview about:
- Approaches to providing care to older adults with insomnia
- Sedative medication prescribing
- Reaction to the intervention materials
- Factors that hinder or support deprescribing of sedative
  3 major themes
-Institutional structure (lack of resources, level of priority)
-Patient characteristics and attitudes (patient dependence, communication)
-Clinician characteristics and attitudes (care burden, alternatives)
Lasserre, 2010 [54] France Quantitative: Cross-sectional survey by questionnaire 350 GPs Questionnaire on:
- Knowledge
- Opinions about prescriptions
- Ways to reduce anxiolytics/hypnotics in older adults
- Barriers
Not reported 82% of GPs knew at least one of the 6 national guidelines on management of people with insomnia and/or anxiety
97.1% of GPs previously felt pressure to renew anxiolytics/hypnotics
90.5% declared that it was possible to reduce or stop treatment for their patients
High level of patient physical and psychological dependence reported
Agreement on ways to reduce anxiolytics/hypnotics:
-Campaign to inform general population: 84%
-Increased access to psychiatrist: 81%
-Reinforcement of physician education: 81%
-Increased access to psychotherapy: 60%
Agreement with identified barriers:
-Patient does not want to stop the treatment: 79%
-Psychotherapy is not refunded: 79%
-Psychotherapy is not accessible: 73%
-No alternative therapy to propose: 70%
-Drug withdrawal more dangerous than benefits of stopping: 58%
-Relatives of the patient refuse drug withdrawal: 38%
-Unaware of drug withdrawal procedure: 8%
Martin, 2017 [55] Canada Mixed methods:
Quantitative study of factors associated with outcome in the EMPOWER study (educational brochure)
261 adults over 65, receiving at least 5 medicines and chronic BZRA prescription Evaluation of the three mechanisms of increasing motivation, capacity and opportunity links with outcome BZRA cessation
Individuals who achieved a dose reduction were classified as intent to deprescribe with failed discontinuation
-Higher perceived necessity score
-Poor health
-Improved knowledge
-Increased concern
-Higher risk perception about BZRA
-Higher self-efficacy to discontinue
   Qualitative interviews 21 adults over 65, receiving at least 5 medicines and chronic BZRA prescription Interviews on:
- Initial reactions to the intervention
- Reasons underlying the decision to taper
- Experience with the tapering process
- Personal interactions with hcps
  Contexts associated with negative outcome (barriers):
-Previous discouragement from physician
-Poor health status
-Unquestioning belief in their physician
-Lack of perception of personal risk
-Reliance on medication for coping/everyday function
-Quality of life focus during end of life
-Discouragement from a physician
-Intolerance to recurrence of symptoms/withdrawal effects
-Loss of confidence to complete the tapering process (post intervention)
Contexts associated with positive outcome (enablers):
-Previous support from physician/positive attitude towards discontinuation
-Stable health status
-Certainty and confidence about tapering (post intervention)
-Positive outlook on ageing
-Perception of increased risk
-Lack of psychological attachment
-Tapering tool provides support
-Supportive HCP
Williams, 2016 [56] Australia Qualitative 17 older adults with at least 2 nocturnal BZRA prescriptions in the last 6 months Interviews on:
-BZRA initiation
-Perception of doctors’ attitudes
-Thoughts on stopping BZRA
-Awareness of alternative treatments
Not reported 3 major themes explored:
-Commencement and continuation of nocturnal BZRA (reasons for use, benefits perceived, previous deprescribing attempts)
-Participants’ knowledge of BZRA and alternative options available (knowledge of adverse effects, treatment alternatives)
-Attitudes to BZRA use and cessation (perceived GP opinion, willingness to deprescribe)
Studies conducted in the hospital setting
Yokoi, 2014 [57] USA Quantitative, retrospective chart review 75 patients on standing BZRA, admitted for at least 14 days Differences at admission between continuers and withdrawers
Factors potentially associated with deprescribing
No BZRA at discharge Withdrawers were less anxious at admission than continuers
Continuers had a better mean orientation score than withdrawers
2 potential enablers for BZRA discontinuation:
-Older age
-Higher antidepressant dose
  1. BZRA benzodiazepine receptor agonist, cRCT cluster randomised control trial, GP general practioner, HCP health care provider, NH nursing home, NHR nursing home resident, TDF theoretical domains framework