Assessing medication burden and polypharmacy: finding the perfect measure

KEYWORDS:

• Polypharmacy
• deprescribing
• older adults

View correction statement:

Erratum

1. Assessing medication burden

Assessing medication burden to evaluate the quality of pharmacotherapy in older adults can be performed using a range of indicators including polypharmacy. Polypharmacy is a common term in the literature, but definitions vary from measuring total number of medications using various cutoffs to the use of high-risk or unnecessary medications [1]. For research purposes, the most common definition identifies polypharmacy as use of five or more medicines. While this approach may be considered arbitrary, there is evidence to support the use of this threshold when assessing if polypharmacy is appropriate or potentially inappropriate (i.e. when harms outweigh the benefits) in clinical practice [2].

Polypharmacy is an international phenomenon, and the prevalence of polypharmacy continues to rise. Polypharmacy is identified in 37.7% of older Australian adults living in the community [3]. Among older Americans aged 65 years and older, from 1999–2000 to 2011–2012, polypharmacy increased from 24% to 39% [4]. Considering the increasing trends in medication use over the last decade, researchers now quantify hyperpolypharmacy exposure, which refers to the use of ≥10 medications. Among older Australian adults living in the community, hyperpolypharmacy prevalence is 4.8% [3] compared to 23.8% among patients admitted to acute general medical hospital services [5].

Over the last 20 years, measures of medication burden have moved from characterizing polypharmacy as counting medications to instruments to standardize assessment of medication burden in older adults, with the ultimate goal to reduce inappropriate polypharmacy and improve patient outcomes. Explicit and implicit instruments, such as the Beers Criteria, STOPP/START criteria, and Medication Appropriateness Index, are common criteria that can be used to identify high-risk medications that suggest reconsideration [1]. Medication burden in older adults can also be assessed using tools that consider pharmacological principles (i.e. dose–response and cumulative effects) and target specific medications such as those with clinically significant anticholinergic effects and sedative effects (i.e. Anticholinergic Drug Score, Anticholinergic Risk Scale, Anticholinergic Cognitive Burden Scale, Sedative Load, and Drug Burden Index). However, there are challenges with comparing medication burden exposure using these tools, specifically anticholinergic tools due to the lack of consensus of what constitutes ‘anticholinergic medication.’ Therefore, tool chosen to assess medication burden will have major impact on the results in pharmacoepidemiological studies assessing the utilization or outcomes of anticholinergic medications. There are new instruments being developed to assess inappropriate medication use among older adults with multiple chronic conditions including dementia. An explicit list of medications that are of questionable benefit among people with advanced dementia has been recently developed [6]. Using this list, it was found that 53.9% of people with advanced dementia living in nursing homes received at least one medication with questionable benefit.

2. Clinical consequences of polypharmacy

To date, evidence regarding clinical consequences of polypharmacy has been mostly established from observational studies. Evidence from systematic review of observational studies suggests that polypharmacy is linked with a range of prescribing and clinical outcomes including drug–drug interactions, medication non-adherence, inappropriate prescribing, adverse drug events, hospitalization, falls, functional decline, and mortality [7]. However, applying this evidence to guide clinical practice and decisions is challenging due to the inconsistent findings reported and methodological limitations observed across studies.

It has been argued that polypharmacy may not be always harmful. In a study of Scottish primary care data of over 180,000 patients aged 20 years and over, increasing number of medications, defined as use of 1–3, 4–6, and 7–9 prescription medications, was not consistently linked with increased hospital admissions among people with multimorbidity [8]. While this may suggest that polypharmacy is appropriate among individuals with multimorbidity, it has also been shown that one-fifth of older adults will receive one medication that will adversely affect coexisting condition [9]. It should be noted that the study by Payne et al. also showed that hyperpolypharmacy was associated with an increased risk of hospitalization across patient groups with any number of comorbidities.

Clearly considering polypharmacy outside the context of multimorbidity will not progress the field nor will it inform clinical practice. This has been acknowledged in the recent National Institute for Health and Care Excellence guidelines which provide recommendations for management of polypharmacy among people with multimorbidity [10]. Indeed, recent work has been focused on generating evidence about the effectiveness of polypharmacy among older adults with multimorbidity and geriatric syndromes including frailty, falls, and functional impairment. Post hoc analysis of the apixaban for reduction in stroke and other thromboembolic events in atrial fibrillation (ARISTOTLE) trial showed that patients with atrial fibrillation treated with apixaban had consistently lower major bleeding rates compared with warfarin treatment, although the magnitude of benefit decreased with the increasing polypharmacy exposure [11]. In community-dwelling older men with ischemic heart disease, greater adherence to optimal therapy was associated with lower risk of institutionalization and mortality, stratified according to presence of geriatric syndromes [12]. In older women with ischemic heart disease, use of optimal medical therapy (including antiplatelet, beta-blocker, renin–angiotensin system blocker, and statin therapy) was associated with increased risk of falls, particularly among frail women, with no statistically significant gain in cardiovascular health [13]. In another study of older adults with multimorbidity, the associations between guideline-recommended medication and survival varied according to coexisting conditions and treatment combinations [14].

3. Advances in observational studies investigating the safety of polypharmacy

Pharmacoepidemiological studies assessing the safety of polypharmacy are increasingly employing novel statistical approaches to more accurately estimate the impact of medications in older adults. One such method is a longitudinal extension of the average attributable fraction, which provides additive attributable fractions accounting for both temporal ordering and time-varying conditions, medications, and other patient characteristics [14,15]. Using this method, it was found that two cardiovascular conditions were attributable to mortality, and all three classes of cardiovascular medications indicated to treat these conditions significantly attributed to survival [14]. In another study, multistate modeling approach was used for the first time to estimate how medication use contributes to transitioning from pre-frail to frail status and subsequent death. The study found that each additional medication was associated with a 22% greater risk of transitioning from the robust state to death [16]. Moreover, every unit increase in exposure to anticholinergic and sedative burden, measured using the Drug Burden Index tool, was associated with a 73% greater risk of transitioning from the robust state to the pre-frail state and a 2.75 times greater risk of transitioning from the robust state to death. Observational studies determining effects of cumulative medication burden are essential to inform the gap between clinical trials and real world and to guide prescribing decisions among adults with multimorbidity and geriatric syndromes.

4. Evidence to inform strategies to reduce polypharmacy

Recent efforts have been focused toward informing evidence base on medication withdrawal or deprescribing. Deprescribing can be referred to as a process of withdrawing inappropriate medications, supervised by a health-care professional with the goal of managing polypharmacy and improving patient outcomes [17]. To date, the success of deprescribing interventions to reduce the medication burden is at best mixed. The outcomes of deprescribing are inconsistent and vary by setting and on the intervention being evaluated [18]. A Cochrane review of interventions to improve the use of appropriate polypharmacy found beneficial effects in reducing inappropriate prescribing and medication-related problems; however, no benefits were observed in terms of clinical outcomes [19]. A structured, multidisciplinary approach including medication reconciliation, medication review conducted by a pharmacist, or use of assessment tools to detect medications known to increase the risk of adverse events may minimize potentially inappropriate prescribing. Moreover, an integrated approach taking into account patient perspectives may results in more successful deprescribing interventions.

There is now emerging evidence that deprescribing strategies targeting specific populations and medication classes may improve outcomes for patients. While the evidence to support deprescribing process is growing, withdrawing medications is often found to be difficult by health professionals. The algorithm to guide deprescribing process to reduce inappropriate polypharmacy in clinical practice has been proposed by the Australian Deprescribing Network [20]. Newly formed Canadian Deprescribing Network is focusing efforts on developing deprescribing guidelines for specific medications. Moreover, computerized clinical decision support systems can help health professionals in the process of evaluating the appropriateness of prescribing, reducing inappropriate polypharmacy, and optimizing health outcomes. The importance of shared decision-making and patient preferences in guiding deprescribing process has been recently highlighted, although implementation in the clinical practice may be challenging [21]. There is a clinical need for future studies to investigate the extent, feasibility, benefits, and risks of deprescribing interventions and to better support health professionals in the process of reducing inappropriate polypharmacy in older adults.

5. Conclusion

Polypharmacy continues to rise, and it poses significant burden to older adults. To ensure appropriate international comparisons, future studies should define polypharmacy more consistently, specific to the population and setting studied. Importantly, clinically relevant cutoffs for polypharmacy must be measured and interpreted according to the clinical context, multimorbidity, patient preferences and goals of care. While the international consensus in relation to assessing medication burden is being sought, efforts should be focused to establish clinical and policy approaches at redressing too much medicine through informing the evidence base on deprescribing.

Declaration of interest

D Gnjidic is supported by the Bridging Support Fellowship, University of Sydney. H Allore is supported by U.S. National Institute of Aging at the National Institutes of Health (R01 AG047891, R21 AG045148, P30 AG021342). The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

Additional information

Funding

This work was supported by the University of Sydney and National Institutes of Health.