Proactive Deprescribing Among Older Adults With Polypharmacy: Barriers and Enablers

Study Design, Setting, and Participant Selection

From March 2023 to March 2024, we conducted a qualitative study with outpatients aged 65 years or older who visited either a primary care group practice or a community hospital in Kanagawa, Japan, and had prescriptions for at least 5 regular medications. Patients were recruited in person by research staff not involved in the direct clinical care of the patients. The study was approved by the Institutional Ethical Committee of St Marianna University School of Medicine (No. 5593; date of approval: June 1, 2022).

Data Collection

A research coordinator or the principal investigator (K.I.) assessed the eligibility of patients who visited the study sites and obtained written consent. Before the interview, we collected participants’ age, sex, activities of daily living, number of regular medications, and number of potentially inappropriate medications (PIMs) based on the 2023 Beers Criteria of the American Geriatrics Society.¹³ For medications, we included only prescribed oral medications with a documented prescription duration of 28 days or longer.

Two authors previously unknown to the patients (K.I. and R.M.) conducted semistructured in-depth interviews in Japanese, following an interview guide specifically developed for this study (Supplemental Appendix). The guide was based on previous studies^7,9,10 and on research team discussions regarding relevant views and experiences of patients: (1) their thoughts on their health problems and prescriptions, (2) their experience with deprescribing interventions and their degree of involvement in the decision making, and (3) how the experience of deprescribing has affected their subsequent life, health, and perspectives.

We conducted individual interviews in private spaces at the research sites. The semistructured interviews were recorded and lasted approximately 20 to 30 minutes, with a mean (SD) duration of 22.6 (11.0) minutes. Verbatim transcripts were prepared and personally identifiable information was removed.

Analysis

The qualitative analysis in this study applied a deductive-inductive approach following methods used by Hahn et al³ and Heser et al.⁵ Thematic analysis was conducted following the framework of Braun and Clarke,¹⁴ which provides a systematic process for identifying, analyzing, and reporting patterns within data. The structure of the codebook, which guided the thematic analysis, included categories, themes (divided into enablers and barriers), and subthemes to provide clarity and organization. The interview transcripts were independently read and coded by 2 authors (K.I. and R.M.), focusing on salience, frequency, and elaboration of themes. The same authors collaboratively developed the codebook in Japanese through iterative discussions after each round of interviews. During initial codebook development, concordance between these 2 authors was assessed, yielding a rate of 78.3%. Disagreements among coders were discussed to reach consensus, and any discrepancies were resolved through group discussions.

After consensus on the codebook was reached, a native Japanese speaker with advanced proficiency in English (K.I.) translated the themes, subthemes, and quotations into English. The translated content was reviewed by the research team. Enrollment continued until theoretical saturation was reached, consistent with the emphasis of LaDonna et al¹⁵ on theoretical sufficiency as a marker of rigor in qualitative research, which we ensured by defining a focused aim, conducting iterative analysis, and refining codes through ongoing discussions.

Through inductive conceptualization, participants were categorized according to their medication valuation (positive, negative, or indifferent), decision-making preference (proactive or passive), and openness to deprescribing (open or cautious), partly referring to the typologies developed by Crutzen et al⁶ and Weir et al,⁸ which classify patients based on their attitudes and behavior toward medications. Two authors (K.I. and R.M.) independently conducted this categorization, and consensus was reached through discussion among the research team. Another author (S.M.A.), who has expertise in anthropology and public health, repeatedly reviewed the study process and the conceptual model for triangulation.

Category 1. Medication Valuation

Category 2. Decision-Making Preference

Category 3. Openness to Deprescribing

Typology

On the basis of the main themes identified from the interviews, we focused on patients’ medication valuation, decision-making preference, and openness to deprescribing, which resulted in 5 typologies partly related to those of Weir et al⁸ and Crutzen et al⁶: indifferent (3 patients, 15%), satisfied and risk-averse (2 patients, 10%), compliant (6 patients, 30%), fearful but passive (4 patients, 20%), and proactive (5 patients, 25%) (Figure 1 and Supplemental Table 1).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

Patient Typology Based on Attitudes Toward Medication, Willingness to Deprescribe, and Proactiveness to Initiate a Conversation About Deprescribing

Note: Number of prescriptions is the mean (SD) number per patient.

^a All proactive patients had a negative medication valuation and an openness to deprescribing.

These 5 patient types represent a synthesis of our findings on enablers and barriers. Although they draw partly on the classifications by Weir et al⁸ and Crutzen et al,⁶ the inclusion of a proactive type is an original contribution to the understanding of patient perspectives on deprescribing. All patients who were categorized as proactive placed a negative value on medication, suggesting that unfavorable perspectives on medication may be an essential trigger of patients’ proactive attitudes toward deprescribing, and also had an openness to deprescribing.

The most frequent patient type was compliant, characterized by positive attitudes toward medications but also a willingness to deprescribe if it was recommended by their prescriber (Figure 1). Although numbers of patients were too small for meaningful comparisons, the mean (SD) number of prescriptions per patient was numerically higher in the compliant group, at 9.8 (3.7), and the indifferent group, at 9.0 (2.6), and lower in the satisfied and risk-averse group, at 5.5 (0.7), and the proactive group, at 6.4 (1.1).

Findings in Context

Previous studies have suggested a potential association between patient proactiveness and views of the value of medication.^6,12 In our study, all patients categorized as proactive had a negative opinion about medications. Although further studies are required to examine the associations, we hypothesize that a patient’s medication valuation may play a vital role in being proactive when it comes to decision making about deprescribing.

A perceived lack of capability to understand and participate in decision making is a known barrier to deprescribing.¹⁶ Individuals who perceived themselves as lacking decision-making capabilities, partly because of older age and multiple comorbidities, may have unrealistic expectations for medication adherence as a self-promoting behavior. Motivational interviewing, a collaborative communication style, could help address these challenges by fostering a patient’s self-efficacy and supporting informed decision making.¹⁷ Evidence shows that motivational interviewing enhances behavioral change by building confidence in taking actionable steps.¹⁸ On the basis of our findings, we hypothesize that motivational interviewing could encourage proactiveness in patients hesitant to initiate deprescribing discussions. Future research may explore the potential of motivational interviewing–based tools to strengthen a patient’s confidence and support a more patient-centered approach to deprescribing.

Positive medication valuation and satisfaction with the existing regimen were common among some patients, reflecting their comfort with prescribed medications and their trust in the current treatment plan. Antonovsky¹⁹ introduced the concepts of sense of coherence and health resources within the framework of salutogenesis. These concepts involve behaviors that form “the self” within daily routines. It is possible that satisfaction with the status quo may reflect the older adult’s sense of coherence through being adherent to prescribed medications. It is therefore likely that there is a group of older adults for whom uniform deprescribing could impact their well-being by bringing about changes in their sense of coherence.

Our results are consistent with those of prior work that emphasized the importance of the health care professional initiating the deprescribing conversation.² Notably, even patients who are positive about their medications or reluctant to ask their prescribers to deprescribe are often willing to stop a medication if the prescriber recommends doing so. This receptiveness suggests that the prescriber should raise the topic even if the patient seems content with the current medication.

Compared with the previous typologies defined by Weir et al⁸ and Crutzen et al,⁶ our typology explicitly focuses on patient proactiveness in deprescribing decision making, extending beyond the willingness to deprescribe, which typically requires a recommendation from health care professionals. We believe that proactiveness toward deprescribing, as highlighted in our study, represents an important factor that has not been sufficiently explored. For instance, patients categorized as proactive may benefit from shared decision-making models that provide more control over medication choices, whereas patients categorized as satisfied and risk-averse or as compliant may require more prescriber-led conversations to explore deprescribing options. Future studies should focus on gaining a deeper understanding of the processes that foster proactiveness and develop interventions that effectively promote it.

Strengths and Limitations

Our results can inform deprescribing frameworks applicable across diverse health care settings. By using strategies to ensure rigor, such as triangulation across different levels—including multiple investigators, a combination of inductive and deductive methods, and a sample with representation of both clinic and hospital patients—we aimed to enhance the trustworthiness of our findings. Furthermore, we included patients with negative perspectives on deprescribing, allowing for greater variability in attitudes and abilities and thereby supporting the credibility of our interpretations.

Study limitations include the following. We used 5 or more regular medications as an inclusion criterion; hence, the perspectives of adults aged ≥65 years who are proactive in avoiding polypharmacy may not be reflected. By recruiting from both primary care and hospital settings, we aimed to enhance the diversity of perspectives; however, cultural factors unique to Japan, such as high trust in health care professionals and the societal emphasis on medication adherence, may have influenced patients’ attitudes toward deprescribing. These cultural variables should be considered when interpreting our findings and assessing their potential applicability to other regions or health care systems.

Conclusions

This study reveals that older adults’ readiness for deprescribing is influenced by their medication valuation, decision-making preferences, and openness to change, and combinations of these attributes can be used to define 5 distinct patient types. Our findings underscore the importance of tailoring deprescribing interventions to the individual patient’s perspective and level of proactiveness. Future research should investigate how patient categorization according to this typology impacts long-term health outcomes and medication adherence after deprescribing interventions. Additionally, studies examining strategies to enhance patient education and self-efficacy among those less open to deprescribing could further inform best practices for reducing polypharmacy in older adults.