Right Message and Right Time: Maximizing Effect of Patient-Directed Education to Promote Deprescribing

The prevalence of and harms associated with polypharmacy and potentially inappropriate medications (PIMs) in older adults have been well explicated. Benzodiazepine receptor antagonists have significant risks for older adults and are categorized as “potentially inappropriate” on the 2023 Beers Criteria, leading them to be a frequent target of medication safety efforts [1]. Deprescribing, an intervention aimed at optimizing medication use, is defined as the supervised cessation or dose reduction of medication. Common implementation strategies to increase deprescribing include system-level changes, clinician-directed, or patient-directed [2-4]. Patient-directed deprescribing strategies are specifically designed to engage and educate patients as active participants in and initiators of decision-making, rather than relying on the clinician or healthcare system to spark change. Such strategies can include written materials, verbal counseling, or digital interfaces. In contrast, clinician-directed strategies focus on changing prescribing behavior, often via guidelines, electronic health record alerts, or academic detailing. The seminal EMPOWER study of benzodiazepine deprescribing among community dwelling older adults presented an effective, low-tech, highly adaptable model of a patient-directed deprescribing strategy and garnered significant attention to more widely adapt and disseminate. It is unsurprising that the low-tech simplicity of patient-directed materials is alluring, especially in the context of complex health systems and difficulty changing prescriber behavior. Further, deprescribing tightly aligns with the expansion of Age Friendly Health Systems (AFHS) and the centrality of “what matters most” to patients [5]. This approach also echoes the American Geriatrics Society's Guiding Principles for the Care of Older Adults with Multimorbidity, which underscores the importance of incorporating patient goals and preferences into shared decision-making [6]. In the years since EMPOWER, there have been numerous studies of patient-directed strategies to increase deprescribing of benzodiazepines and other medication classes.

McEvoy and colleagues sought to assess the impact of patient-directed, non-pharmacological strategies to deprescribe benzodiazepine receptor antagonists (i.e., benzodiazepines and z-drugs) used to treat insomnia in adults aged 65 years and older and in people living with cognitive impairment [7]. They identified 17 reports from 16 studies; given the heterogeneity of the interventions and the settings, the results were summarized by narrative review. Papers were broadly categorized as either solely patient-directed education or multi-component (e.g., addition of prescriber-directed activities), with further division by whether the patient education was written, verbal, or both. Of the solely patient-directed educational studies, six featured written materials only, and five included written plus verbal. These studies were generally successful at reducing benzodiazepine use compared to usual care, albeit with a wide range of cessation rates (14%–72%) and with no apparent differences between either written or verbal education compared to a combination of the two. Six additional studies included patient education as a part of a multi-component strategy to deprescribe benzodiazepines. These were also typically effective but again with extremely diverse rates of deprescribing (9%–100%).

Health-related information provided by clinicians is subject to low or inaccurate recall by patients and retention is inversely related to information volume [8, 9]. Thus, considerable attention has been given to improving patient recall of healthcare communication. Studies comparing communication modality (verbal vs. written) show mixed results, often favoring the combination of verbal information and written materials [10]. In this review by McEvoy and colleagues, there appears to be little difference in efficacy of written materials only compared to written materials augmented with verbal components. In settings where clinicians often feel overtasked and team-based resources and options may be restricted, it raises the question of whether there is sufficient incremental benefit when written materials, as a low cost and low-tech strategy, may adequately move the needle on reducing use of risky medications like benzodiazepines. The EMPOWER brochures exemplify this type of approach—offering a patient-directed strategy that has demonstrated effectiveness across various care settings [4, 11, 12].

The wide range of success across included studies warrants deeper examination of the factors that led to variable findings to best understand which aspects should be retained for future implementation. One such aspect might be the timing of the educational delivery and alignment of care interaction with a patient “primed” for a deprescribing discussion. For community dwelling older adults, when is the optimal timing? It could be coordinated with a scheduled outpatient visit, facilitating discussions with trusted clinicians that incorporate principles of patient-centered care (e.g., shared decision-making, alignment with what “matters most”) [11]. Acute hospitalization and post-acute care may represent another valuable window of opportunity, as patients are often experiencing changes in health status and may be reflecting on what matters most to them [12-14]. During hospital stays, patients are frequently asked about their preferences for life-sustaining treatments—though the quality of these conversations and the extent of shared decision-making can vary. When done well, these discussions should include an assessment of goals of care and what matters most, in line with AFHS principles [13, 14]. If medications are not aligned with goals, that misalignment can be used to reinforce PIM deprescribing [15]. Just as acute care can be a “teachable moment” for other behavioral changes (e.g., tobacco cessation, dietary habits), the inpatient setting may be an untapped opportunity to educate patients about risks of their medication(s). The GABA-WHY trial demonstrated the success of a patient-directed deprescribing intervention targeting gabapentinoid prescribing among hospitalized older adults with a median length of stay of approximately 10 days [12]. Research is needed to define the most effective timing and setting for delivering patient-directed deprescribing interventions in real-world care environment.

The concept of optimal timing can also refer to timing along the prescribing continuum. Deprescribing is inherently an action that occurs after someone is taking a medication. However, this does not automatically mean that strategies to implement deprescribing can only occur at this same point in the continuum. It is plausible to integrate deprescribing concepts and patient education earlier in the prescribing continuum, especially given findings that continuing a medication is not viewed as negatively as initiating a medication [16]. In addition, patients may have misperceptions about longer duration of illness indicating greater need for aggressive treatment whereas guidelines for management of diseases like diabetes support less aggressive management as risk benefit ratios shift [17]. Capitalizing on these concepts, a study of hypothetical proton pump inhibitor (PPI) use found that including warnings about long-term use at the time the PPI is started is associated with subsequent increased interest in discontinuing it [18]. Whether these findings translate to real-world patient scenarios requires evidence. While it is possible that patient-directed education related to deprescribing can have a role when initiating medications, it will need to be balanced with long-standing efforts to improve medication adherence to not cause unintended consequences.

Of note, despite the intention of the authors to assess benzodiazepine deprescribing among those with dementia, only one study included in this review approached this topic. The absence of evidence is particularly concerning given the high prevalence of benzodiazepine use among older adults with dementia and their potential for heightened vulnerability to adverse effects. The singular trial included in the review by McEvoy and colleagues was conducted among those with mild cognitive impairment—a population different than those with more advanced dementia when considering educational interventions, expected comprehension, and potential participation in care decisions. Perhaps, then, it is unsurprising that a post hoc analysis from the EMPOWER trial found no significant differences in the effectiveness of written education for older adults with mild cognitive impairment compared to those without any cognitive impairment—a finding warranting further investigation [19]. The D-PRESCRIBE-AD trial also used patient-directed written materials and was unsuccessful in increasing deprescribing rates for targeted PIMs (antipsychotics, sedative-hypnotics, anticholinergics) among those with dementia [20]. While acknowledging the difficulty including patients with dementia in research studies, dedicated studies of patient-directed strategies to increase deprescribing in this population are much needed. It is it possible that materials may need to be adapted to enhance understanding, additional individuals may need to be included (e.g., care partners), or other features require modification. This is particularly salient given the significant need and opportunity to align medications with what matters most among patients with dementia.

The review by McEvoy and colleagues highlights both the promise and limitations of the current literature as it relates to patient-directed implementation strategies to increase benzodiazepine deprescribing in older adults. Written educational materials, with their relatively low cost and implementation burden, appear to be as effective as more resource-intensive multi-component strategies, though with substantial variability in outcomes. Future research should address gaps identified by this review, including factors influencing variability in deprescribing rates between interventions and implementation strategies, utility of pre-emptive prescribing warnings, and effective approaches for patients with dementia. Clarity about what drives the substantial variability in deprescribing outcomes can allow us to better align prescribing practices with AFHS and the prioritization of what matters most to patients.

A.M.L. and K.M.Z. equally contributed to conceptualization, visualization, writing the original draft, review and editing, and approval of the final submission.

The authors declare no conflicts of interest.

This publication is linked to a related article by McEvoy et al. To view this article, visit https://doi.org/10.1111/jgs.19512.