Effectiveness of Interventions to Improve the Anticholinergic Prescribing Practice in Older Adults: A Systematic Review

Background: Pharmacotherapy in older adults is one of the most challenging aspects of patient care. Older people are prone to drug-related problems such as adverse effects, ineffectiveness, underdosage, overdosage, and drug interactions. Anticholinergic medications are associated with poor outcomes in older patients, and there is no specific intervention strategy for reducing drug burden from anticholinergic activity medications. Little is known about the effectiveness of current interventions that may likely improve the anticholinergic prescribing practice in older adults. Aims: This review seeks to document all types of interventions aiming to reduce anticholinergic prescribing among older adults and assess the current evidence and quality of existing single and combined interventions. Methods: We systematically searched MEDLINE, Embase, Cochrane Central Register of Controlled Trials, CINAHL, and PsycINFO from January 1990 to August 2021. Only studies that examined the effect of interventions in older people focused on improving compliance with anticholinergic prescribing guidelines with quantifiable data were included. The primary outcome of interest was to find the effectiveness of interventions that enhance the anticholinergic prescribing practice in older adults. Results: We screened 3168 records and ended up in 23 studies that met the inclusion criteria. We found only single-component interventions to reduce anticholinergic prescribing errors in older people. Pharmacists implemented interventions without collaboration in nearly half of the studies (n = 11). Medication review (43%) and education provision (26%) to healthcare practitioners were the most common interventions. Sixteen studies (70%) reported significant reductions in anticholinergic prescribing errors, whereas seven studies (30%) showed no significant effect. Conclusion: This systematic review suggests that healthcare practitioner-oriented interventions have the potential to reduce the occurrence of anticholinergic prescribing errors in older people. Interventions were primarily effective in reducing the burden of anticholinergic medications and assisting with deprescribing anticholinergic medications in older adults.


Introduction
Prescribing medications among older adults is recognised as a challenging task and an essential practice that needs to be continuously monitored, assessed, and refined accordingly. Moreover, it is based on understanding clinical pharmacology principles, knowledge about medicines, and particularly the experience and empirical knowledge of the prescribers [1,2]. Clinicians face several challenges while prescribing medications among older adults, and the prescribing of potentially inappropriate medications (PIMs) for this age group is prevalent [3]. The available epidemiological data show that up to 20% of older patients in outpatient settings and 59% of hospitalised older patients consume at least one PIM [4][5][6][7][8]. Adverse effects in older people due to inappropriate prescribing are prevalent, leading to increased hospital admissions and mortality [9].
Medications that possess anticholinergic activity are a class of PIMs widely prescribed for various clinical conditions in older adults [10,11]. Older people are particularly vulnerable to the adverse effects from medicines with anticholinergic-type effects [12,13]. Most medications commonly prescribed to older people are not routinely recognised as having anticholinergic activity, and empirically, clinicians prescribe these medicines based on their anticipated therapeutic benefits while overlooking the risk of cumulative anticholinergic burden [14][15][16]. Anticholinergic burden refers to the cumulative effect of taking multiple medications with anticholinergic activity [17,18]. There is no gold standard approach available to quantify and determine whether an acceptable range of anticholinergic drug burden exists in older adults [19,20]. The central adverse effects of anticholinergic medications are attributed to the excess blocking of cholinergic receptors within the central nervous system (CNS) [16]. The commonly reported central adverse effects are cognitive impairment, headache, reduced cognitive function, anxiety, and behavioural disturbances [16]. The common peripheral adverse effects of anticholinergic medications are hyperthermia, reduced saliva and tear production, urinary retention, constipation, and tachycardia [16].
Anticholinergic medications are associated with poor outcomes in older patients, but there is no specific intervention strategy for reducing anticholinergic drug exposure [21]. There is little evidence that medication review could be a promising strategy in reducing the drug burden in older people [22,23]. Medical practitioner-led and pharmacist-led medication reviews have earlier been reported as a standard practice for reducing anticholinergic drug exposure [24,25]. Pharmacist-led medication review has recently been found to be ineffective among older patients of the Northern Netherlands [25]. A few metaanalyses have also reported the lack of effectiveness of different types of medication reviews on mortality and hospitalisation outcomes [26][27][28]. Multidisciplinary strategies such as patient-centred, pharmacist-physician intervention are also recognised as promising for improving medication use in older patients at risk [29]. Another intervention strategy, i.e., the SÄKLÄK project, had some effects on the PIMs prescription and reduced potential medication-related problems [30]. The SÄKLÄK project is a multi-professional intervention model to improve medication use in older people [30], and it consists of self-assessment using a questionnaire, peer-reviewed by experienced healthcare professionals, feedback report provided by experienced healthcare professionals, and an improvement plan [30].
Interventions to improve prescribing practice more generally have been the subject of many studies and are frequently targeted according to the type of error [31,32]. It is crucial to explore which interventions have effectively changed prescribing practices and optimised patient outcomes while minimising healthcare costs. However, little is known about the effectiveness of existing interventions at improving the anticholinergic prescribing practice for older adults. Hence, this review seeks to document all types of interventions aiming to reduce anticholinergic prescribing errors among older adults and assess the evidence of existing single and combined interventions.

Methods
The Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guideline was applied to report the findings of this systematic review [33].

Data Sources and Search Strategy
The following databases were examined between January 1990 and August 2021: Ovid MEDLINE, Ovid EMBASE, Ovid PsycINFO, and the Cochrane Central Register of Controlled Trials (CENTRAL). A comprehensive electronic search was performed using appropriate keywords on anticholinergics, older people, and interventions to retrieve the relevant studies. The search was limited to the English language and humans. A detailed MEDLINE search strategy is presented in Supplementary Table S1. Citation analysis was performed in Google Scholar and Web of Science to track the prospective citing of references of the selected articles.

Study Screening and Selection
The title, abstract, and full text of each potentially relevant article were independently screened by two authors (M.S. and S.A.) for eligibility of inclusion in this review. Any discrepancies were resolved by a third author (A.A.), and decisions were made by consensus.

Inclusion Criteria
The primary outcome of interest was single and multicomponent interventions that improve anticholinergic prescribing practice or reduce adverse drug events due to the consumption of anticholinergic medications. Single-component intervention consists of only one intervention activity, such as medication review [34]. Multicomponent intervention refers to the combination of various components in a single intervention [35], such as medication review and the provision of education [34]. All interventions (e.g., medication review, educational detailing visits for physicians, nurses, and aides, pocket-sized educational cards along with clinical vignettes, educational internet site, and detailing session with physicians) performed by any healthcare professional targeting participants of either sex, mean age ≥65 years, and admitted to any healthcare setting were included. We included pre/post or experimental studies that employed a control group.

Exclusion Criteria
We excluded the following studies: review articles, case reports, and case series. We also excluded studies that were conducted in languages other than English.

Data Extraction and Synthesis
Two reviewers (M.S. and S.A.) independently reviewed and extracted the data from the eligible studies according to a standardised format based on variables of interest, such as the study population, study design and duration, mean age, major findings, and intervention characteristics (type of intervention and implementation). The study selection process is illustrated in Figure 1.

Quality Assessment
The quality of the included studies was critically appraised. The Cochrane Risk of Bias tool [36] was used to assess the methodological quality of the randomised controlled trials (RCTs). The Newcastle-Ottawa scale was used to assess the quality of the non-RCTs [37], which is based on three domains: the selection of study groups, comparability of cohorts and assessment of outcome (cohort studies), or comparability of case and controls and ascertainment of exposure (case-control studies). The thresholds for categorising and interpreting the Newcastle-Ottawa scale domains were described in Supplementary Table S2. The Cochrane Risk of Bias tool results for included RCTs were described in Supplementary  Table S3. Studies were not excluded based on the risk of bias or quality assessment.

Quality Assessment
The quality of the included studies was critically appraised. The Cochrane Risk of Bias tool [36] was used to assess the methodological quality of the randomised controlled trials (RCTs). The Newcastle-Ottawa scale was used to assess the quality of the non-RCTs [37], which is based on three domains: the selection of study groups, comparability of cohorts and assessment of outcome (cohort studies), or comparability of case and controls and ascertainment of exposure (case-control studies). The thresholds for categorising and interpreting the Newcastle-Ottawa scale domains were described in Supplementary Table  S2. The Cochrane Risk of Bias tool results for included RCTs were described in Supplementary Table S3. Studies were not excluded based on the risk of bias or quality assessment.

Results
The primary electronic search identified a total of 3168 studies from the five databases. Using EndNote X9 (Thomson Reuters), we eliminated 350 duplicate studies, and the remaining 2818 studies were examined to determine their relevance for inclusion. Of

Results
The primary electronic search identified a total of 3168 studies from the five databases. Using EndNote X9 (Thomson Reuters), we eliminated 350 duplicate studies, and the remaining 2818 studies were examined to determine their relevance for inclusion. Of those, only 70 were found to be eligible for full-text analysis. Subsequently, 47 studies were excluded as they failed to meet the predefined inclusion criteria. No potential studies were identified from the citation analysis. Finally, a total of 23 studies that investigated the effectiveness of anticholinergic prescribing practice in older adults were included in this review ( Figure 1). Table 1 provides the qualitative summary of the included studies, mainly showcasing the type of interventions, and Table 2 illustrates an overview of the quantitative summary of the studies based on study design, setting, sample size, study duration and follow-up, outcome measure (control/pre and intervention/post), significant association (+ or −), and statistical tests.
No significant impact on the medication risks between the intervention and the control group. The per-protocol analysis indicated a tendency for effectiveness, particularly in optimising central nervous system medication use.

Medication review
Pharmacists reviewed the medications and detected drug-related problems using the Drug Burden Index (DBI) tool. Their recommendations were communicated to the physician via telephone, weekly meetings, and email. Further review was conducted at the weekly meeting between physician and pharmacist.
Statistically significant differences were found between pre-and post-intervention in NPI at admission, drug-related problems, MAI criteria (interactions, dosage and duplication), and mean (SD) DBI score.
Lenander et al., 2018, Sweden [51] Cross-sectional Medication Review Clinical Pharmacist led medication review to assess the prevalence of DRPs and recommendations to discontinue, followed by team-based discussions with general practitioners (GPs) and nurses It shows that the medication reviews decreased the use of potentially inappropriate medication.

Medication Review
Medication review was conducted for ACB in patients at the time of admission and discharge 21.1% of patients had their ACB reduced. There is considerable scope for improvement of prescribing practices in older people. Physicians in the primary care settings could communicate with pharmacists employing a shared EHR. As part of a quality improvement project, a pharmacist-led EHR-based medication therapy recommendation service was implemented at 2 DHS medical clinics to reduce the anticholinergic burden High recommendation acceptance rates were achieved using objective anticholinergic risk assessment and algorithm-driven medication therapy recommendations. Physicians receiving the CDSS issued more discontinuation orders of definite anticholinergics, but the results were not statistically significant. Results suggest that human interaction may play an important role in accepting recommendations aimed at improving the care of hospitalised older adults with CI. Cluster RCT Medication review The study intervention included a letter and phone call to GPs, using DBI to prompt them to consider dose reduction or cessation of anticholinergic and sedative medications.
At follow-up, a DBI change was observed in 16 participants. DBI decreased in 12 participants, 6 (19%) in the control group, and 6 (32%) in the intervention group. Educational visits used academic detailing to discuss prescribing of highly anticholinergic antidepressants in elderly people.
The rate of starting anticholinergic antidepressants in the elderly reduced 26% (in the individual intervention) and 45% (in the group intervention) The use of less anticholinergic antidepressants increased by 40% and 29%, respectively
Healthcare practitioner-initiated education mainly consisted of professional components, such as academic detailing sessions for physicians [47,54,55], evidence-based training programs to improve dispensing [39], interactive training sessions for nurses [45], and mailing of intervention letters to the physicians [38]. In three studies [21,29,30], healthcare practitioners also performed interventions such as targeted patient-centred, pharmacistphysician team medication therapy management (MTM) intervention, SÄKLÄK project, and multidisciplinary medication review in collaborations with other healthcare practitioners. A context-aware pharmaceutical analysis tool was tested in France to automatically compare prescriptions recorded in computerised patient files against the main consensual guidelines [52]. Another study tested the clinical decision support system to discontinue orders of definite anticholinergic medications for hospitalised patients with cognitive impairment [41]. Similarly, a study tested targeted patient-centred pharmacist-physician team MTM intervention to reduce the consumption of inappropriate anticholinergic medications in older patients [29]. In Italy, researchers tested the INTERcheck computerised prescription support system to optimise drug prescriptions and minimise the occurrence of adverse drug reactions [53].
Healthcare practitioner-oriented interventions that reported a significant reduction in anticholinergic prescribing errors included: medication review, education provision to healthcare practitioners, pharmacist-led electronic health record-based population health initiative and anticholinergic risk scale service, targeted patient-centred, pharmacistphysician team MTM intervention, context-aware pharmaceutical analysis tool, and SÄK-LÄK project. Healthcare practitioner-oriented interventions were most effective in reducing ACB [21,29,40,54], DBI [22,42,48,50], and discontinuation or reduction of anticholinergic medications [30,38,39,43,46,51,52,55]. Hernandez et al. 2020 reported a decline in DBI from 1.38 (control group) to 1.08 (intervention group) [50]. Another study reported a reduction in ACB score from 1.08 (control group) to 0.89 (intervention group) [40]. A retrospective study by McLarin et al. [43] in Australia found a reduction in the mean scores of anticholinergic medications from 3.73 to 3.02 after implementing medication review.

Discussion
This is believed to be the first systematic review assessing the effectiveness of interventions to reduce anticholinergic prescribing errors in adults aged 65 and above. Previous reviews primarily evaluated the studies of pharmacist-oriented interventions on medication prescribing and the association between anticholinergic drug burden and mortality in older people [17,56,57]. We did not conduct a meta-analysis because of the methodological heterogeneity between the study designs, anticholinergic prescribing errors, types of interventions, study duration, and follow-up period. Given the high prevalence of inappropriate prescribing and polypharmacy in older people aged 65 and over, interventions to reduce anticholinergic prescribing errors in this cohort are of considerable importance. This systematic review identified 23 studies reporting interventions to reduce anticholinergic prescribing errors in older people. The interventions were mainly provided by the pharmacists using a patient-centred approach. Many studies (19 out of 23) successfully reduced the incidence of anticholinergic prescribing errors in older people. Evidence related to the pharmacist-led interventions in many studies suggests that pharmacists play a vital role in the care of older people, thus improving medication safety across the continuum of care.
In this study, medication review and education provision to the healthcare practitioners were the most common elements in many interventions. Medication review is a structured evaluation of patients' pharmacotherapy to optimise drug use and reduce the occurrence of drug-related problems [58]. Similarly, medication review is recognised as an important healthcare practitioner-oriented intervention for reducing anticholinergic prescribing errors in older people [59]. Likewise, older people benefit mostly from medication reviews as this cohort is more susceptible to adverse drug effects [60,61]. The efficacy of medication review in reducing anticholinergic prescribing errors was reported by eight studies in this review [15,21,22,43,44,46,50,62]. Previous studies inform the significant effects of structured medication review on medication prescriptions and older adults' quality of life [63][64][65].
Another intervention, such as the provision of education to the healthcare practitioners, was tested in eight studies, but only five studies reported the effectiveness of this intervention in reducing anticholinergic prescribing errors in older people [38,49,54,55,66]. Evidence informs that the healthcare practitioner-oriented educational intervention effectively reduces prescribing errors in older people [57,67]. The provision of education reduces the use of healthcare resources, including emergency department presentations and hospital admissions [68]. Implementing healthcare practitioner-led educational interventions encourages prescribers to change prescription practices, thus improving prescribers' clinical practice [69]. An education intervention provides precise knowledge about prescribing in older adults, medication-related errors, and prevention strategies for reducing medication-related errors [69]. This review also showed that interventions such as INTERcheck, SÄKLÄK intervention model, targeted MTM intervention, context-aware pharmaceutical analysis tool, and CDSS alert were not successful in reducing anticholinergic prescribing errors in older adults [29,41,[51][52][53].

Implications for Clinical Practice and Future Research
Medications with anticholinergic activity are frequently prescribed in older people due to their numerous clinical benefits; however, these medications are also associated with poor clinical outcomes [70]. Implementing healthcare practitioner-oriented interventions can reduce the occurrence of anticholinergic prescribing errors in older people. This review's findings inform that healthcare practitioner-oriented interventions appear to improve medication safety in older people based on observed reductions in anticholinergic prescribing errors, particularly when the provision of care involves a medication review and an education for the physicians prescribing anticholinergic medications. The prescribing competency can be optimised through educational interventions [71], thus reducing the occurrence of anticholinergic prescribing errors. The safe and effective prescribing of medications is a challenge in older people who frequently experience multiple long-term conditions and complex polypharmacy [72]. Due to an increasing challenge to physicians when prescribing and the complexity of medication regimens taken by older people, there is a need to embed prescribing competency framework in clinical practice [72]. The prescribing competency framework engages prescribers in different stages of prescribing, such as information gathering, clinical decision making, communication, monitoring, and review [72].
The current quantification methods for anticholinergic burden tend to streamline the complexity of pharmacological mechanisms in geriatric risk assessment in older adults. However, there is no universally accepted quantification method available to estimate anticholinergic drug burden, and it is difficult to compare the study findings from distinct methods [16]. Existing tools derived from expert consensus limit the quantification of anticholinergic burden as they do not take into consideration the dose and the CNS distribution of drugs [15]. A recent review showed that the ratings of anticholinergic activity in the expert opinion scales were inconsistent [73]. Moreover, the estimation of central cognitive effects by measuring in vitro serum assay of medications with known anticholinergic activity as a composite peripheral measure still remains unclear [16]. The lack of a gold standard method for anticholinergic quantification might have a direct or indirect impact on the interpretation of the effect size of the study interventions (e.g., reduction in anticholinergic burden). In this review, many studies (69%) were either cross-sectional or nonrandomised and included a single-component intervention. Therefore, there is a need to conduct future randomised multicomponent intervention trials for evaluating the true impact of healthcare practitioner-oriented interventions on anticholinergic prescribing errors in older people.

Strengths and Limitations
This study was comprehensive in that the electronic search, conducted in four important databases, attempted to identify the complete existing body of evidence of the effectiveness of healthcare-oriented interventions aimed at reducing anticholinergic prescribing errors in older people. This is the first review that found 17 different types of healthcare-oriented interventions, and their impact on anticholinergic prescribing errors. A limitation of this study included the absence of meta-analysis and the estimation of the effect size. It was mainly due to the heterogeneity of included studies. We also excluded studies published in languages other than English, which may have introduced a language bias. However, we performed citation tracking and hand-searching of all included studies to minimise the influence of factors (e.g., inconsistent terminology or wrong indexing) that may affect the keyword-based search.

Conclusions
This systematic review suggests that healthcare practitioner-oriented interventions have the potential to reduce the occurrence of anticholinergic prescribing errors in older people. Medication review and the provision of education to the prescribers were the most common approaches to reducing anticholinergic prescribing errors in older people. Healthcare practitioner-oriented interventions were mostly effective in reducing the burden of anticholinergic medications and facilitating the discontinuation of anticholinergic medications in older people. In the future, there is also the need to ascertain how often the healthcare practitioner performs interventions that may reduce anticholinergic prescribing errors in older people.

Conflicts of Interest:
The authors declare no conflict of interest.