Features of successful medication review and deprescribing interventions for fall prevention in residential aged care facilities: an intervention component analysis of an updated systematic review

Abstract

Background

Medication review and deprescribing are common fall prevention strategies, as falls risk increasing drugs and polypharmacy are potentially modifiable falls risk factors. We sought to understand why some trials in residential aged care facilities (RACFs) testing medication review/deprescribing reduced falls whilst others did not. We used intervention component analysis (ICA) to develop a theory of the key feature(s) from the trialists’ perspective.

Methods

Randomised controlled trials (RCTs) evaluating medication review/deprescribing as a single falls prevention intervention in RACFs, were identified via a Cochrane review and additional database searches to 1 April 2025. ICA was conducted with inductive thematic analysis of the Discussion sections of trial publications. Alignment between themes/subthemes and trial outcomes were examined.

Results

Thirteen trials from six countries were included. Trialists perceived that medication review/deprescribing interventions should include a tool (i.e. algorithm/list of medications) to facilitate decisions between clinicians on the appropriateness of resident’s prescriptions, with at least one prescriber from the residents’ usual care team involved in decisions and implementation of medication changes, reviewing medications together at least 6 monthly. These features were present in 100% (3/3) successful trials and 20% (2/10) unsuccessful trials.

Conclusions

ICA indicated using a tool as a guide amongst clinicians, at least six monthly and including at least one prescriber from the usual care team, could be an important combination of features to improve intervention success. This approach which aligns with recent World Falls Guidelines should be tested in future RCTs of medication optimisation for people living in RACFs.

Key Points

• Medication review and deprescribing are common fall prevention strategies.

• A tool to facilitate medication change decisions between clinicians could be important for successfully reducing falls in RACFs.

• Medication change decisions should involve the resident’s usual prescriber.

• The tool should consider appropriateness of all medications.

• Clinicians should consider an individual resident’s prescribing pattern, at least 6 monthly.

Introduction

Fall risk increasing drugs (FRIDs) and polypharmacy are known and preventable falls risk factors [1]. Amongst community-dwelling older adults, the daily consumption of one FRID or two to five prescription medications have been associated with a 90 to 96% increased falls risk over 3 years, compared to older adults who had no prescription medications [2]. Similarly, amongst older adults in residential aged care facilities (RACFs), polypharmacy or the consumption of FRIDs, such as opioids, anticholinergic, psychotropic or anti-dementia drugs has been associated with falls [3].

Medication review for older adults to consider the appropriateness of prescriptions is a common fall prevention strategy, featured as one subdomain of multifactorial fall prevention trials in RACFs [4]. Hence, medication review/deprescribing are strategies recommended by recent guidelines [5]. The World Guidelines for falls prevention and management amongst older adults recommends a regular routine medication review with appropriate deprescribing of FRIDs by using a validated and structured tool to identify FRIDs, as part of a multifactorial intervention [5, 6]. Suitable tools identified by the guideline included the Screening Tool for Older Persons’ Prescriptions and Screening Tool for Alert to Right Treatment (STOPP/START), Screening Tool of Older Persons Prescriptions in older adults with high fall risk (STOPPFall), Beer’s criteria, Fit for the Aged (FORTA), Turkish Inappropriate medication use in the elderly (TIME), Review of potentially inappropriate medications prescribing in seniors (REMEDI[e]S) and Web-based Meds75 + guide [6]. Medication review is acknowledged as an important component of multifactorial interventions both in the community and in RACFs [7]. However, trials of medication review/deprescribing as single interventions amongst RACF residents are often ineffective at preventing falls [8, 9]. Yet, some individual trials have been effective [10, 11]. Implementation of medication review interventions are complex. Various tools can be used by different health professions, in the highly complex RACF setting, where staffing arrangements can vary. Acknowledging the need for further randomised controlled trials (RCTs), intervention component analysis (ICA) of previous RCTs could assist the design of future RCTs. ICA is a method that enables the exploration of complex interventions to identify combinations of important features for trial success [12]. ICA examines the perspectives of trialists from publications to identify a theory of possible drivers of successful trials [12]. This approach has been useful in other areas of falls prevention research [4, 13]. This study aims to identify a theory of key features of single intervention fall prevention medication review/deprescribing trials in RACFs that may improve falls prevention.

Methods

This study uses ICA; a synthesis method of qualitative information in systematic reviews [14].

Search strategy and selection process

Trials included in the 2018 Cochrane Collaboration review [8] plus records from additional search updates of CENTRAL, MEDLINE, Embase and CINAHL Plus databases from 1 September 2017 to 1 April 2025 were screened to identify eligible trials. Trial records and conference abstracts were not systematically searched or included. Records retrieved from the updated search were imported into Endnote x9 (Clarivate Analytics, PA, USA) for duplicate removal. The remaining records were imported into Covidence (Veritas Health Innovation, Melbourne Australia) for independent screening by two reviewers. Any discrepancies were discussed with a third reviewer when necessary.

Inclusion criteria

RCTs testing a medication review/deprescribing intervention as a single intervention compared to usual care in RACFs, published in English and reporting data suitable to determine the ratio of the rate of falls or risk of falling were included.

Whilst medication review and deprescribing are not synonyms, medication review was defined as the practice of evaluating an individual’s medication/s with the aim of optimising medication use, which can also include the practice of deprescribing (i.e. reducing the medication dose, tapering the dose to cease the medication or removal of medication/s). Therefore, trials that appeared to be providing similar medication change interventions, aiming to reduce falls risk as determined by two reviewers were included.

Exclusion criteria

Trials with an intervention that did not directly provide medication review/deprescribing for residents, for example interventions that provided staff education [15, 16] or tested an automatic geriatric risk assessment report [17] were excluded. Trials that provided a multifactorial intervention or multiple interventions that include medication review/deprescribing intervention as one of two or more components according to the Prevention of Falls Network Europe taxonomy [18, 19], were excluded. In addition, trials that did not report falls data suitable for meta-analysis were excluded from this review [20, 21].

Data extraction

Study characteristics (i.e. number randomised, mean age, percent female, intervention description, control description and falls outcomes used) were extracted independently by two reviewers. Risk of bias ratings and falls outcomes from eligible trials published in 2018 Cochrane review were extracted [8]. The same Cochrane Collaboration risk of bias 1 tool [22] was used via Covidence to assess selection bias, performance bias, detection bias, attrition bias, reporting bias, as well as bias associated with the method of ascertaining falls, baseline imbalances and other bias, for new trials as described previously [8].

Outcomes measures

Raw data or statistics on the rate or number of falls and the number of participants falling one or more times was extracted. The rate ratio (ratio of the number of falls per unit time) or risk ratio (ratio of the number of people falling one or more times) between the intervention and control arms and 95% confidence interval (CI) was determined according to the methods described in Supplementary materials (Appendix 1), following the methods reported in the 2018 Cochrane Collaboration review [8].

Intervention component analysis

Research papers published in English associated with the included trials were imported into NVivo 20 (QSR International Pty Ltd, Release 1.6.1) for inductive line-by-line coding of trialist perspectives on critical features associated with success or failure of their trial at reducing falls [12].

First, the discussion and conclusions sections of the primary trial paper were read and inductively coded, followed by reports associated with the trial (i.e. programme evaluation paper, implementation paper and intervention development paper) to code additional information. Initial coding was conducted by one author with a random sample checked by a second author for agreement. One author then completed initial coding and grouped similar initial codes considering the relationship between them following the process of inductive thematic analysis of Braun and Clarke [23].

Second, groupings of initial codes and relationships were checked and discussed with a second author to establish the themes (i.e. main features) and subthemes (i.e. components of main features). Codes not associated with a theme or subtheme but instead represented thoughts that were only amongst one or two trials, were presented as minor views.

Third, preliminary themes, subthemes and minor views were discussed amongst all reviewers including experts on medication review/deprescribing, fall prevention and RACF research (e.g. physicians, pharmacists, fall prevention researchers), to consider the interpretation of the themes, subthemes and minor views for logic and commonality of thoughts.

Finally, the presence and absence of themes and subthemes along with the success or failure of the intervention in preventing falls were considered [12]. Two reviewers used standard definitions derived from the ICA to independently determine the presence or absence of identified trial features, with discussion with a third author as required. Where the presence or absence of the identified features was not clear from the published records, trialists were contacted for additional information. The success of the trials was determined using the rate ratio or risk ratio for falls. The rate ratio was used in preference to the risk ratio for falling when available, as the rate of falls appears to be more sensitive to change [8]. The success of the intervention was defined as reducing falls (when the ratio of the rate of falls or risk of falling point estimate was ≤0.8). The failure of the intervention was defined as no clear effect on falls (ratio of the rate of falls or risk of falls point estimate >0.8 to <1.2) or increased falls (ratio of the rate of falls point estimate ≥1.2).

Results

Thirteen trials across six high-income countries met the inclusion criteria (Figure 1). Seven trials provided a medication review intervention, amongst residents in RACFs. One of these seven trials focused on medication review for first time residents transitioning from hospital to RACF [24]. The remaining six trials provided a medication deprescribing intervention to residents in RACFs, focusing on psychoactives [25], site selected drug classes [26] and multiple drug classes based on patient care [27–30]. Residents were predominantly females (61% to 77% of the trial population) in their 80s, living in a RACF in a high-income country (Table 1). Based on the falls rate or risk ratios, three trials were successful at reducing falls [10, 11, 28] and the remaining 10 trials were unsuccessful at reducing falls, with two trials reporting data suggesting increased falls [25, 30] (Table 1). Risk of bias of the included trials is considered in Supplementary materials, Appendix 2.

Figure 1.

Study Selection Process [31]

^* Excluding studies providing education to staff only or testing a risk assessment and not directly providing medication review/deprescribing intervention for residents as the intervention.

Table 1.

Study characteristics.

Author, year primary paper, other related records	Study design	Country	Residents N	Age in years mean (SD) or median (range)	Female %	Intervention	Control	Intervention dose	Timepoint that falls is measured
Cateau 2021a [26] Mena 2023 [32]	RCT (cluster)	Switzerland	Resident numbers NR*	NR	NR	Interprofessional Quality Circle-Deprescribing Module (QC-DeMo). Deprescribing based on pharmacist selected drug classes. Pharmacists of the nursing homes participated in a half-day education session. Pharmacists selected classes based on drug use in the nursing home, organised and facilitated the session, with nurses and physician, formalised a deprescribing consensus via two quality circle sessions to enacted consensus with choice to hold supplementary sessions during the year as needed.	Usual care	One medication withdrawal plan at one timepoint	12 months
Cateau 2021b [29]	RCT (individual)	Switzerland	T: 62 I: 32 C: 30	T: 85.5^ I: 87 (80–91) C: 84 (78–88)	T: 63%^ I: 52%, (n = 16) C: 74% (n = 20)	Pharmacists took part in a 3-day postgraduate education session on performing medication review, not part of pharmacy curriculum. Individual deprescribing-focused medication review, performed by the pharmacists using standardised form for performing review and documenting treatment modification plan. Single treatment modification plan in collaboration with nurses and physicians.	Usual care	One medication review at one timepoint	4 months
Crotty 2004 [24]	RCT (individual)	Australia	T: 110 I: 56 C: 54	T: 82.7 (6.4) I: 82 (80.2–83.7) C: 83.4 (81.7–85.1)	T: 61% I: 58.9% C: 63%	Pharmacist coordinated the transition for patients transferring from hospital to a RACF for the first time. This included medication management transfer summaries (hospital), medication review (accredited community pharmacists) and case conferences with physicians and pharmacists. All regular and as needed medications prescribed as of the date of hospital discharge (baseline) and 8 weeks after discharge (follow up) were included in MAI assessment to determine appropriateness of medications.	Usual hospital discharge process	One medication review within 10 to 14 days of transfer	8 weeks
Curtin, 2020 [28]	RCT (individual)	Ireland	T: 130 I: 65 C:65	T: 85.1 (5.7) I: 84.5 (5.6) C: 85.7 (5.9)	T: 62%^ I: 65% C: 58%	Medication withdrawal plan based on the STOPPFrail criteria was devised by the research physician was communicated directly to one of the participant’s attending physicians and documented in the patient’s medical record. The attending physician judged whether to accept the drug withdrawal plan and implement the recommended changes. One review was conducted.	Usual care	One medication withdrawal plan at one timepoint	3 months
Desborough 2020 [33]	RCT (cluster)	UK	T: 826 I: 381 C: 445	T: 87 I: 88.4 (6.5) C: 86.0 (8.5)	T: 76.0%^ I: 79.5%^ C: 72.8%^	Multiprofessional medication review. Multi-professional medication review (MPMR): meeting between clinical pharmacist and pharmacy technician, care home staff and GP(s). Review conducted at baseline and 6 months.	Usual care	MPMR at baseline and 6 months	12 months
Etherton-Beer 2023 [27]	RCT (individual)	Australia	T: 303 I1: 102 I2: 101 C: 100	T: 85.0 (SD 7.5) I1: 85.8 (7.1) I2: 84.8 (7.7) C: 85.0 (7.2)	T: 76 I1: 75 I2: 75 C: 77	Deprescribing intervention. Medication review for withdrawal in accordance with structural deprescribing. Two research pharmacists independently reviewed all medications taken regularly or at least weekly to generate a list of medications targeted for withdrawal according to an algorithm. Structured plan for order of withdrawal of medications and tapering. GP given medication withdrawal plan and asked to confirm agreement prior to randomisation. Blinding by encapsulating medications targeted for deprescribing. One arm blinded by encapsulation, one arm unblinded.	Blind control (medications encapsulated but continued)	One medication withdrawal plan at one timepoint	12 months
Frankenthal 2014 [10] Frankenthal 2015 [34] Lavan 2017 [35]	RCT (individual)	Israel	T: 359 I: 183 C: 176	T: 82.7 (8.7) I: NR C: NR	T: 67.0 I: 70.5 C: 62.5	Medication review by pharmacist with Screening Tool of Older Persons PIPs/Screening Tool to Alert doctors to Right Treatment (STOPP/START). Pharmacist made recommendations to chief physician who decided whether to implement changes.	No interventional recommendations made by pharmacist to chief physician	Medication review at baseline, 6 months and 12 months	12 months
Holland 2023 [36] Birt 2021 [37]	RCT (cluster)	UK	T 882 I: 454 C: 428	T: 85.3 (7.7) I: 85.1 (7.7) C:85.4 (7.6)	T: 69.6 I: 72% C: 67%	Medication review. Pharmacist independent prescribers (PIPs) visited care homes to do medication reviews and create pharmaceutical care plans. Usual care provided general support for improving home processes and staff training. PIPs received 6-week training including STOPP/START criteria for medication review. No set method of communication and external to team.	Usual care (could range from 3, 6 or 12 monthly visits by primary care-based pharmacist for medication review)	Medication review weekly over 6 months	6 months
Patterson 2010 [25]	RCT (cluster)	UK	T: 334 I: 173 C: 161	T: 82.7 (8.4) I: 82.6 (8.4) C: 82.9 (8.4)	T: 73 I: 72.3 C: 73.4	Pharmacist review of psychoactive medications monthly for 12 months working with prescribers (GPs) to improve prescribing of these drugs. Monthly visits for 1. Assessment of residents’ pharmaceutical care needs 2. Medication review 3. Preparation of a pharmaceutical care plan shared between the relevant healthcare personnel 4. Pharmacist intervention and direct communication with the prescriber	Usual care	Medication review monthly over 12 months	12 months
Potter 2016 [30]	RCT (individual)	Australia	T: 95 I: 47 C: 48	T: 84.3 (SD 6.9) I: 84 (6) C: 84 (8)	T: 52 I: 55 C: 48	Deprescribing by GP and geriatrician who was also a clinical pharmacologist. Medicine withdrawal plan guided by an evidence-based algorithm, amended to reflect changes requested by participant, next-of-kin, or GP, was implemented over several months. GP reviewed participants weekly during deprescribing.	Usual care	Medication withdrawal plan, twice weekly monitoring of any dose reductions until target drugs are ceased	12 months
Roughead 2022 [38]	RCT (individual)	Australia	T: 282 I: 136 C: 146	T: 86 (7.4) I: NR C: NR	T: 68 I: NR C: NR	Medication review. Pharmacists were trained, undertook medication review, met with resident and care staff, conducted assessments of cognition, movement behaviour and hand grip strength and liaised with resident’s doctors.	Usual care: (Residential medication management review every 12 to 24 months)	Medication review, every 8 weeks over 12 months	12 months
Wouters 2017[39]	RCT (cluster)	The Netherlands	Individuals: T: 426 I: 233 C: 193	T: 83.5^ I: 83.7 (9.5) C: 83.2 (8.9)	T: 68^ I: 65 C: 71	One multidisciplinary multi-step medication review including considering the STOPP/START criteria and Beer’s criteria was carried out by the patient’s treating elderly care physician and a pharmacist, with training. Received brief training before performing it. Pharmacist completed three steps: Patient perspectives collected on a questionnaire, drug reviewing by pharmacist using criteria, multidisciplinary meeting, with physician completing fourth step of executing actions.	Usual care: (Medication safety monitoring; ad hoc medication reviews when clinical indicated; but no standardised multidisciplinary medication reviews as per intervention)	One medication review at one time point	4 months
Zermansky 2006 [11] Lowe 2000 [40]	RCT (individual)	United Kingdom	T: 661 I: 331 C: 330	T: 85 (80–90) I: 85.3 (81–90) C: 84.9 (80–90)	T: 76.7%^ I: 77.3%^ C: 76.1%^	Medication review by pharmacist with carer and patient. Passed on a written proforma to GP for acceptance and implementation where GP acceptance was signified by ticking a box on the proforma.	Usual care (by GP)	One medication review within 28 days of randomisation	6 months

C: control; CF: care facilities; GP: general practitioner; I: intervention; MAI: Medication appropriateness index, MPMR: multiprofessional medication review, NFU: no follow-up; NHS: National Health Service (UK); NR: not reported; RACF: residential aged care facility; RaR: rate ratio (rate of falls); RR: risk ratio (number of fallers); T: total; UK: United Kingdom; ^ reviewer calculated * residents were from 55 RACF with 17 RACF randomised to intervention and 29 randomised to control.

Intervention component analysis

Eighteen published records associated with the 13 included trials were available for thematic analysis (Table 1). From the perspective of trialists’, there was one key theme for optimal implementation of medication review/deprescribing interventions to reduce falls in RACFs. Trialists perceived that medication review/deprescribing interventions should include the use of a tool to guide and facilitate clinicians to consider the appropriateness of the patient’s prescribing pattern. The tool should consider multiple classes of medications associated with potentially inappropriate prescriptions (PIPs) or potential prescription omissions (PPOs). A prescriber who is part of the resident’s usual healthcare team should be one of the clinicians that considers the tool as they implement the final medication changes (Table 2).

Table 2a.

Thematic analysis of trialist perspectives on key features of medication review/deprescribing intervention for reducing falls in RACFs.

Theme	Subthemes	Key informal evidence
Using a tool (covering the individual’s prescribing pattern), as a guide for clinicians to consider the appropriateness of the patient’s prescribing pattern with usual staff involved in implementation (at least 6 monthly)	Considering patients at the centre of the process when changing the prescribing pattern (adjusting, starting, or reducing medications) (40 codes, 10 trials)	‘Moreover, optimal prescribing is based on improving prescribing patterns and not on decreasing the number of medications. The gradual decrease in PIPs/PPOs in the intervention group demonstrated the improvement in prescribing in the current study.’ [25] ‘Quality deprescribing, as with good prescribing, requires the patient to be at the centre of the process. The risks of each medicine need to be weighed against the expected benefits in this specific person at this specific point in their life, taking into account their preferences and expectations, their likely prognosis, their co-morbidities, their symptoms, their other medicines, and the wishes and expectations of their family or carers. This is not a straightforward process and there is no simple list, guideline, or algorithm that will make it so.’ [30] ‘Adjusting or starting medication that improves mobility (such as anti-Parkinson’s medication) is another’. [11]
	Tool to facilitate decisions between clinicians (43 codes, 9 trials)	‘..this study did not use the STOPP criteria to aid clinical decisions,. . . . . .. . .. Recent evidence suggests using prescribing appropriateness tools may be able to demonstrate patient benefit’ [33] ‘The criteria are not designed to replace clinical judgement, but rather to assist clinicians with medication reviews and assessment of treatment goals in this specific patient cohort.’ [35] ‘Other structured deprescribing methods have recently been evaluated in very frail older people using a randomized controlled trial design, and they also reported statistically significant reductions in potentially inappropriate prescriptions. Potter et al used an implicit (i.e., judgment-based) algorithm that requires the user to answer a series of questions about each drug in the patient’s regimen; Wouters et al evaluated the Multidisciplinary Multistep Medication Review. Both methods are patient centered and comprehensive but limited by a requirement for resource-intensive processes that may hinder their integration into routine clinical practice. STOPPFrail overcomes these limitations by virtue of its conciseness and high interrater reliability between users of different disciplines and professional grades.’ [28] In this study, the driving factor for success was the high acceptance rate of STOPP (82.4%) and START (92.6%) recommendations by the chief physician. The implementation of STOPP/START criteria in chronic geriatric facilities does not guarantee success. The most dominant cause of success is the adherence of the attending physician to the recommendations based on the STOPP/START criteria after the assessment of harm versus benefit in a patient-specific context. [10] ‘Assessment of appropriate prescribing in older patients is difficult, particularly when it is necessary to assess psychoactive drug use. . .. We found, however, that the MAI was a useful way of operationalizing an assessment of evidence-based prescribing.’[24]
	Working together to decide and act, especially with a prescriber from the usual care team (16 codes, 7 trials)	‘A change in care home managers often led to the leadership in the care home not “buying” into the intervention. There was strong evidence that where care home staff were unsure of the intention of the intervention there was reluctance to fully engage and there were occasional attempts to block pharmacist review.’ ‘. . .. importance of time to build relationships and the management style of the care home manager as being important for successful intervention implementation’ [37] ‘Clinicians who promote change in the diffusion of information among staff have been shown to promote changes in practice behaviours.’ [24] ‘GPs would prefer to work with people who they knew and trusted.’ [37] ‘STOPPFrail criteria were developed in the university affiliated with these hospitals, and this may have influenced the readiness of attending physicians to implement the deprescribing recommendations.’ [28]
	Review as a team on individual basis, at least 6 monthly (7 codes, 4 trials)	‘In contrast, there was a constant increase in the number of medications in the control group (from a mean of 8.2 ~ 3 to a mean of 8.9 ~ 3) and no significant change in PIPs and PPOs. This suggests a need for a medication review using the STOPP/START criteria every 6 months at least.’ [10] ‘may also be that periodic review every 6 months is not sufficient as benefits of the review are lost as a resident’s health status changes and any prescribing changes may be reversed.’ [33]

Tool to facilitate decisions between clinicians including usual staff and used at least 6 monthly

From the trialists’ perspective, the way the tools were used within research trials in RACFs affected the success or failure of the intervention to reduce falls in RACFs (Table 2a). Trialists’ perceived four important considerations: (i) the tool should be used as a guide [10, 11, 24, 25, 28, 33, 36, 38, 39], (ii) any changes to the prescribing pattern (deprescribing, starting or adjusting) should consider the patient [10, 11, 25, 26, 28–30, 33, 36, 39], (iii) clinicians need to work together with involvement from the usual staff to decide and implement any agreed changes [10, 24, 27–29, 37, 39] and (iv) the tool should be used at least 6 monthly [10, 11, 25, 33] (Table 2a).

Whilst not all trials used a tool to identify FRIDs, multiple trialists [33, 35] mentioned that tools only aid clinical decision making and are not designed to replace clinical judgement (Table 2a). Therefore, it would be inappropriate to use the tool (i.e. a list or algorithm) to implement medication changes without considering the patient-specific context. The patient’s overall health goals as well as the patient’s risks and benefits are areas to consider when making clinical medication judgements. Changing the patient’s prescribing pattern amongst older adults was described as not only about deprescribing (i.e. planned and supervised dose reduction or ceasing medications that cause harm or are no longer beneficial), but also about starting medications that are beneficial, if required.

Trialists also perceived the importance of working with another clinician in making clinical decisions, particularly with members of the patient’s existing health care team (Table 2a). One trial described the different clinical expertise that pharmacist and general practitioners offer on medication knowledge, the patient’s context, preferences and goals to consider the patient in a more holistic manner, when changing the prescribing pattern. Seven trials [10, 26–28, 33, 38, 39] mentioned the need to involve and establish a relationship with the treating physician and RACF.

Conducting medication review on an individual basis at least every 6 months was also mentioned amongst four trials [10, 11, 25, 33] and could be an important consideration as medication appropriateness may change with the resident’s health status (Table 2a).

Minor views

Two trials discussed the importance of medication review/deprescribing as having an impact on recurrent falls [10, 26] and as part of multifactorial interventions [33, 39] (Table 2b).

Table 2b.

Minor views.

Other minor views	Informal evidence
Reductions in medications impact recurrent falls (9 codes, 2 trials)	The current study intervention revealed a strong association between PIPs and recurrent falls. The reduction in PIPs did not significantly change the prevalence of falls, but it significantly reduced the number of recurrent falls in the intervention group. [10]
Multifactorial intervention required (2 codes, 2 trials)	‘multi-factorial in nature and require complex interventions as they are also related to the ergonomics of the care home environment and physical frailty of the care home population, consequently it may not be appropriate to expect a single intervention focussed on medication to have a significant impact on falls.’ [33]

Examination of trialists’ perspectives in relation to trial outcomes

Based on falls rate ratio or risk ratio, three trials were considered successful at reducing falls [24, 27, 28] and the remaining 10 trials unsuccessful (Table 3). Standard definitons and evidence used to allocate trials according to presence and absence of the theme and subthemes (i.e. tool used as guide, working together and at least 6 monthly) are summarised in the Supplementary Materials Appendix 3.

Table 3.

Presence of themes and subthemes amongst falls prevention trials of medication review/deprescribing in RACFs.

#	Trial year	Theme	Subthemes			RaR/RR (95%CI)
		Tool used as a guide for clinicians (including ≥ 1clinician from the usual care team) to work together implementing medication changes at least 6 monthly (combination of the subthemes)~	Tool used as guide to consider the resident’s prescribing pattern ~	Working together with prescriber from usual care team~	At least one review 6 monthly~
Successful trials (RaR/RR ≤ 0.80)
1	Frankenthal 2014 [10]	Yes	Yes	Yes	Yes	RaR 0.61 (0.48–0.78)^
2	Zermansky 2006 [11]	Yes	Yes	Yes	Yes	RaR 0.62 (0.53–0.72)
3	Curtin 2020 [28]	Yes	Yes	Yes	Yes*	RaR 0.68 (0.40–1.15)^
Unsuccessful trials (RaR/RR > 0.80)
4	Holland 2023 [36]	No	Yes	No	Yes	RaR 0.91 (0.67–1.25)
5	Cateau 2021b (IDel) [29]	No	No	Yes	Yes*	RaR 0.94 (0.44–2.02)^
6	Wouters 2017 [39]	Yes	Yes	Yes	Yes*	RaR 0.94 (0.50–1.76)
7	Cateau 2021a (QC-DeMo) [26]	No	No	Yes	No	RaR 0.98 (0.88–1.09)^
8	Desborough 2020 [33]	No	No	Yes	Yes	RaR 1.01 (0.74–1.38)
9	Roughhead 2022 [38]	No	No	No	Yes	RaR 1.01 (0.83–1.23)^
10	Etherton-Beer 2023 [27]	No	Yes	Yes	No	RaR 1.06 (0.91–1.24)^
11	Crotty 2004 [24]	Yes	Yes	Yes	Yes*	RR 1.19 (0.71–1.97)
12	Patterson 2010 [25]	No	No	Yes	Yes	RaR 1.43 (1.07–1.92)^
13	Potter 2016 [30]	No	No	Yes	No	RaR 1.67 (1.34–2.07)^

RaR: rate ratio (rate of falls), RR: risk ratio (number of fallers)

*Trial intervention period was less than 6 months, ^ Reviewer calculated, ~ Standard definitions used are in Supplementary Materials Appendix 3.

All (3/3) ‘successful’ trials and 40% (4/10) ‘unsuccessful’ trials, used a tool as a guide to consider the patient in changing the prescribing pattern (Table 3). All (3/3) successful trials and 80% (8/10) unsuccessful trials had at least two clinicians working together to decide upon medication changes (Table 3). All (3/3) successful trials and 70% (7/10) unsuccessful trials used the tool at least 6 monthly or were trial of 6 months or less (Table 1 and Table 3). The combination of all three of these features was present in 100% (3/3) of successful trials and only 20% (2 of 10) unsuccessful trials (Table 3).

Discussion

This analysis has indicated that conducting medication review/deprescribing interventions by involving at least one prescriber from the usual care team, using a tool (a list or algorithm of questions/medications) at least 6 monthly and considering individual resident’s prescribing patterns could be important factors in improving falls reduction in RACFs. This combination of factors was consistently present in trials that showed a tendency to reduce falls but were infrequently present in trials that were less successful. The inclusion of medication review/deprescribing interventions as a key component of successful multifactorial falls prevention trials has raised questions about how to successfully reduce falls following medication review/deprescribing as a single intervention. Whilst there are only a small number of trials showing a reduction in falls included in this analysis, their use in generating a theory to identify a more successful approach is clearly needed. The elements of this theory have been derived from qualitative analysis of the views of trialists of both successful and unsuccessful trials and have face validity as they are consistent with other sources.

The concept of using a tool at least 6 monthly aligns with the recent recommendations in the World Guidelines for Falls Prevention [5] and the European Geriatric Medicine Society position statement [6]. In the current analysis, the type of tool used was a list or algorithm of questions or medications. Some tools used in the included trials clearly considered FRIDs by dedicating a specific section to them (e.g. STOPP/START and Beer’s criteria). Other trials did not include a specific list of FRIDs (e.g. MAI or a deprescribing algorithm). Regardless, using a systematic method of considering appropriateness of medications is likely to have led to recommendations for the deprescribing of some FRIDs (Supplementary Material, Appendix 3). Using a tool as a guide ensures a consistent approach and provides a summary of the clinical reasoning or workflow considered by the health professional conducting the medication review/deprescribing intervention. Consistency and clear reasoning are likely factors that assist the prescriber from the usual care team to participate in the decision and successful implementation of medication review/deprescribing interventions. Interventions that did not use a tool as guide either used a tool to determine medication changes for all residents [26], only included a medication review discussed at multidisciplinary meeting [33], or not discussed [38], or used a tool that only focused on one medication class [25].

A recent realist review of medication review/deprescribing in primary care also emphasised the importance of having a clear process with a ‘predefined workflow plan’ that allowed the usual care prescriber to have the final decision on medication changes. This led to higher acceptance rates of deprescribing recommendations in this setting [41] and provides a potential solution to working with prescribers who may focus on time efficiency [42]. This analysis indicated that having multiple clinicians, with a least one clinician from the usual care team involved in the decision-making process, may improve the chances of successful falls prevention. A multidisciplinary approach is also recommended in the European Geriatric Medicine Society position statement [6]. Whilst a multidisciplinary structure is essential for optimal care of older adults [43], this can be difficult to achieve in settings where prescribers and pharmacists may not be working for the same facility.

In this ICA, there were two trials that included the three elements of the theory that were not successful at reducing falls [24, 39]. Whilst these two trials involved a multidisciplinary team, the relationship between the usual care clinicians and the use of an external pharmacist may have affected the success of the intervention and created difficulty in timely implementation. In a process evaluation of another included study the prescriber did not wish to work with a pharmacist that they did not have an established relationship with [37]. However, more detailed examination of the strength of the relationship between the clinicians in the included trials was not possible as clear information on this was rarely reported. External staffed pharmacists may no longer be a problem in the future in the United Kingdom [44] and Australia [45], where funding models for on-site care home pharmacists have been announced after these two trials were conducted [24, 36]. It is likely that having multidisciplinary same-site clinicians working together is preferable.

Strengths and limitations

Based on ICA methodology, trialist perspectives are not collected from formal research such as a focus group or interviews of trialists and therefore are not equivalent in depth of information. However, the perspectives gathered from discussion sections of published papers nevertheless reflect opinions of trialists and their experiences when implementing interventions. The allocation of presence and absence of themes and subthemes to individual trials is subjective, but bias was minimised by having two reviewers independently allocate trials, providing quotes from published papers as justification. When there was inadequate evidence to make an allocation, reviewers corresponded with trialists to obtain additional information. Four trialists contacted about 3 trials, all responded [46–48].

As this analysis did not consider recurrent falls as an outcome nor include trials of multifactorial interventions, the minor views on these issues were not examined. Any medication review/deprescribing intervention approach that impacts on recurrent fallers will be reflected in the rate of falls outcomes and thus such features are considered indirectly.

In this analysis, a successful trial was defined as a clinically meaningful 20% or greater reduction in the point estimate of the ratio of the rate of falls or risk of falling, irrespective of statistical significance, in order to consider potentially successful trial approaches regardless of study power. One trial classified as successful in this analysis [28], whils having no effect on the risk of falling (risk ratio 0.90, 95% CI 0.48 to 1.69) had a point estimate indicating a possible 32% reduction in the rate of falls (Table 3, rate ratio 0.68, 95% CI 0.40 to 1.15), which is considered highly clinically significant. Thus, whilst not statistically significant, this trial indicated a trend to a reduction in falls for both the rate of falls and risk of falling and was classified as successful in accord with the a priori methods.

Only a small proportion of the trials included in this analysis were classified as successful, but the 13 included trials provide the best available evidence to guide the design of future trials of medication review/deprescribing as a single intervention to prevent falls in RACFs. Importantly, the views of trialists from both successful and unsuccessful trials on likely key trial features are considered in the ICA. There is a possibility that the reduction in falls in one trial classified as successful in this analysis was due to chance as the reduction observed was not statistically significant [28]. Nevertheless, even if this trial were reclassified as unsuccessful, the presented theory of drivers of success, including a combination of three key features, would still be supported as a possible approach to improve successful implementation, as it would be present in 100% (2/2) of successful trials and only 27% (3/11) of unsuccessful trials.

Whilst there is a possibility that there are additional factors beyond this theory that may contribute to trial success, this ICA generates a much-needed theory for further testing. ICA has been successfully applied in other contexts to generate theories that have been further supported by qualitative comparative analysis, showing that the theory can appropriately explain differences in trial outcomes [4].

Conclusion

This ICA indicated that using a tool (i.e. an algorithm of questions or list/checklist of medications to consider), as a guide amongst clinicians (with at least one prescriber responsible for clinical care of residents in RACF), to consider the individual resident’s prescribing pattern could be important features in improving the success of medication review/deprescribing in reducing falls in RACFs. In line with the European Geriatric Medicine Society statement, performing a medication review/deprescribing intervention at least 6 monthly for individual residents also appeared to be important. This approach should be considered in future medication review/deprescribing falls prevention intervention trials in RACF.

Declaration of Conflicts of Interest:

None.

Declaration of Sources of Funding:

The Centre of Research Excellence for Prevention of Falls Injuries (grant number 1198371) and Medical Research Futures Fund (grant number 2018573) funded by the Australian National Health and Medical Research Council provides salary support for authors Dyer and Suen. Sherrington receives salary support from an Australian National Health and Medical Research Council Investigator Grant. The funders had no role in study design, data analysis and the decision to publish this manuscript.

Research Data Transparency and Availability:

Supplementary files contain all available data.