Deprescribing Education vs Usual Care for Patients With Cognitive Impairment and Primary Care Clinicians: The OPTIMIZE Pragmatic Cluster Randomized Trial

Abstract

Background: Individuals with dementia or mild cognitive impairment frequently have multiple chronic conditions (defined as ≥2 chronic medical conditions) and take multiple medications, increasing their risk for adverse outcomes. Deprescribing (reducing or stopping medications for which potential harms outweigh potential benefits) may decrease their risk of adverse outcomes.

Objective: To examine the effectiveness of increasing patient and clinician awareness about the potential to deprescribe unnecessary or risky medications among patients with dementia or mild cognitive impairment.

Design, setting, and participants: This pragmatic, patient-centered, 12-month cluster randomized clinical trial was conducted from April 1, 2019, to March 31, 2020, at 18 primary care clinics in a not-for-profit integrated health care delivery system. The study included 3012 adults aged 65 years or older with dementia or mild cognitive impairment who had 1 or more additional chronic medical conditions and were taking 5 or more long-term medications.

Interventions: An educational brochure and a questionnaire on attitudes toward deprescribing were mailed to patients prior to a primary care visit, clinicians were notified about the mailing, and deprescribing tip sheets were distributed to clinicians at monthly clinic meetings.

Main outcomes and measures: The number of prescribed long-term medications and the percentage of individuals prescribed 1 or more potentially inappropriate medications (PIMs). Analysis was performed on an intention-to-treat basis.

Results: This study comprised 1433 individuals (806 women [56.2%]; mean [SD] age, 80.1 [7.2] years) in 9 intervention clinics and 1579 individuals (874 women [55.4%]; mean [SD] age, 79.9 [7.5] years) in 9 control clinics who met the eligibility criteria. At baseline, both groups were prescribed a similar mean (SD) number of long-term medications (7.0 [2.1] in the intervention group and 7.0 [2.2] in the control group), and a similar proportion of individuals in both groups were taking 1 or more PIMs (437 of 1433 individuals [30.5%] in the intervention group and 467 of 1579 individuals [29.6%] in the control group). At 6 months, the adjusted mean number of long-term medications was similar in the intervention and control groups (6.4 [95% CI, 6.3-6.5] vs 6.5 [95% CI, 6.4-6.6]; P = .14). The estimated percentages of patients in the intervention and control groups taking 1 or more PIMs were similar (17.8% [95% CI, 15.4%-20.5%] vs 20.9% [95% CI, 18.4%-23.6%]; P = .08). In preplanned subgroup analyses, adjusted differences between the intervention and control groups were -0.16 (95% CI, -0.34 to 0.01) for individuals prescribed 7 or more long-term medications at baseline (n = 1434) and -0.03 (95% CI, -0.20 to 0.13) for those prescribed 5 to 6 medications (n = 1578) (P = .28 for interaction; P = .19 for subgroup interaction for PIMs).

Conclusions and relevance: This large-scale educational deprescribing intervention for older adults with cognitive impairment taking 5 or more long-term medications and their primary care clinicians demonstrated small effect sizes and did not significantly reduce the number of long-term medications and PIMs. Such interventions should target older adults taking relatively more medications.

Trial registration: ClinicalTrials.gov Identifier: NCT03984396.

Conflict of interest statement

Conflict of Interest Disclosures: Drs Bayliss, Gleason, Green, Maciejewski, Sheehan, Wolff, and Boyd; Mss Shetterly, Sawyer, and Kraus; and Messrs Norton and Maiyani reported receiving grants from the National Institute on Aging during the conduct of the study. Dr Green reported receiving grants from the National Institute on Aging Impact Collaboratory during the conduct of the study. Dr Reeve reported receiving grants from the National Institutes of Health (subaward to Dr Reeve’s institution) during the conduct of the study and royalties from UpToDate for writing a chapter on deprescribing. Dr Maciejewski reported receiving Veterans Affairs Health Services Research and Development funding and owning Amgen stock due to his spouse’s employment. Dr Boyd reported receiving royalties from UpToDate for writing a chapter on multimorbidity and honoraria from Dynamed for reviewing a chapter on falls outside the submitted work. No other disclosures were reported.

Figures

Figure 1.. Study Cohort Flow Diagram

MCI indicates mild cognitive impairment.

Figure 2.. Differences Between Intervention and Control Subgroups in Long-term Medication Counts at 6 Months

Estimated differences from linear regression models accounting for baseline counts of medications, age, sex, race and ethnicity, and a random clinic effect. Subgroup models added the appropriate subgroup variable and an interaction with study group. Error bars indicate 95% CIs. aPatients taking 7 or more medications vs 5 to 6 medications (P = .28 for interaction). bTwo mailings vs 1 mailing (P = .70 for interaction). cAlzheimer disease or dementia vs mild cognitive impairment (P = .50 for interaction).

Figure 3.. Differences Between Intervention and Control Subgroups in Percentage of Individuals Taking 1 or More Potentially Inappropriate Medications (PIMs) at 6 Months

Estimated differences from logistic regression models accounting for baseline PIM, age, sex, race and ethnicity, and a random clinic effect. Subgroup models added the appropriate subgroup variable and an interaction with study group. Error bars indicate 95% CIs. aPatients taking 7 or more medications vs 5 to 6 medications (P = .19 for interaction). bTwo mailings vs 1 mailing (P = .70 for interaction). cAlzheimer disease or dementia vs mild cognitive impairment (P = .31 for interaction).