Highlights
- Electronic Health Record (EHR) interventions designed with behavioral science principles increased the likelihood of deprescribing by 26% to 40%.
- The precommitment strategy, which asks clinicians to commit to a deprescribing discussion early in the patient encounter, showed the most robust effect.
- Targeted medications included high-risk classes such as benzodiazepines and anticholinergics, known for their association with falls and cognitive decline in the elderly.
- The study provides a evidence-based framework for health systems to utilize clinical informatics to combat the growing burden of polypharmacy.
Background: The Clinical Imperative for Deprescribing
In the management of older adults, the transition from therapeutic benefit to pharmacological burden is often subtle but clinically significant. Potentially inappropriate medications (PIMs) are drugs for which the risk of adverse events outweighs the clinical benefit, particularly when safer alternatives are available. Despite clear evidence linking PIMs—such as benzodiazepines, non-benzodiazepine sedative-hypnotics (Z-drugs), and anticholinergics—to increased risks of falls, fractures, delirium, and cognitive impairment, their use remains prevalent in geriatric populations.The challenge for primary care physicians (PCPs) is not necessarily a lack of knowledge regarding these risks, but rather the clinical inertia and time constraints inherent in modern practice. Deprescribing is a complex process requiring patient counseling, shared decision-making, and careful tapering. While Electronic Health Records (EHRs) have traditionally been used to drive the adoption of evidence-based therapies, their utility in guiding the cessation of therapy has been less explored. This study investigates whether behavioral science-informed EHR prompts can overcome clinical inertia and facilitate the deprescribing process.
Study Design and Methodology
The study was a 3-group parallel cluster-randomized clinical trial conducted within an academic medical center in Massachusetts. A total of 201 primary care physicians were randomized in November 2022 to one of three arms: usual care, a precommitment intervention, or a boostering intervention.The patient population included adults aged 65 years or older who had a scheduled visit with a participating PCP between November 2022 and March 2024. Eligible patients were those currently prescribed high-risk medications, specifically at least 90 pills of benzodiazepines or sedative-hypnotics, or at least two different anticholinergic medications within the previous 180 days.
The Behavioral Interventions
The trial tested two distinct applications of behavioral science:
1. Precommitment Intervention
This arm utilized the psychological principle that individuals are more likely to follow through on an action if they have made a prior commitment. When the patient’s record was opened during their first visit, the EHR sent a message asking the PCP to initiate a deprescribing discussion. This was followed by a second reminder during the patient’s subsequent visit to encourage the execution of the deprescribing plan.
2. Boostering Intervention
This arm focused on maintaining clinician attention through repeated cues. PCPs received a notification encouraging deprescribing at the first visit, followed by an ‘In-Basket’ reminder four weeks later to serve as a secondary nudge.The primary outcome was defined as the deprescribing (discontinuation or tapering) of at least one target medication between the first patient visit and the end of the follow-up period. Statistical analysis utilized generalized estimating equations to account for clustering at the physician level, with Holm-Bonferroni corrections applied for multiple comparisons.
Key Findings: Driving Clinical Action Through Informatics
The trial followed 1,146 participants with a mean age of 73.6 years. The results indicated a significant increase in deprescribing activity in both intervention arms compared to the control group.
Deprescribing Rates and Relative Risk
In the usual care group, 26.8% of patients (106 individuals) had at least one medication deprescribed. This baseline reflects the natural rate of medication adjustment in a standard academic primary care setting.In the precommitment group, the rate rose to 36.8% (145 individuals). This represented a 40% increased likelihood of deprescribing compared to usual care (Relative Risk [RR], 1.40; 95% CI, 1.14-1.73), with an absolute difference of 10.4%.In the boostering group, 34.3% of patients (122 individuals) underwent deprescribing. This was a 26% increase over usual care (RR, 1.26; 95% CI, 1.01-1.57), with an absolute difference of 6.5%.
Statistical Significance
The primary analysis confirmed that both behavioral interventions were statistically superior to usual care. The precommitment strategy appeared to have a slightly stronger effect size, suggesting that the timing and nature of the ‘nudge’—asking for a commitment before the clinical encounter fully unfolds—may be particularly effective in altering prescribing habits.
Safety and Clinical Outcomes
Monitoring safety during deprescribing is paramount, as tapering certain medications, particularly benzodiazepines, can lead to withdrawal symptoms if not managed correctly. Reassuringly, no serious adverse events related to the deprescribing process were reported through the formal adverse event reporting system.Manual chart reviews were conducted to assess mortality rates during the follow-up period. Death rates were 1.4% in the precommitment group, 3.9% in the boostering group, and 1.8% in the usual care group. While the numerical increase in the boostering group was noted, the study was not powered to detect differences in mortality, and these figures likely reflect the underlying health status of the elderly cohort rather than a direct consequence of the intervention.
Expert Commentary: Behavioral Nudges vs. Alert Fatigue
The success of this trial lies in its nuanced approach to EHR integration. Traditional ‘pop-up’ alerts are often criticized for contributing to clinician burnout and ‘alert fatigue,’ leading many physicians to dismiss them reflexively. By using behavioral science principles like precommitment, the researchers moved beyond simple warnings toward a structured decision-making framework.The precommitment intervention is particularly interesting from a cognitive psychology perspective. By prompting the physician at the start of the visit, the intervention influences the ‘agenda-setting’ phase of the clinical encounter. This makes it more likely that the physician will allocate time to discuss medication risks with the patient, rather than addressing it as an afterthought at the end of a busy session.One limitation noted by health policy experts is the single-center nature of the study. Academic centers may have different baseline prescribing cultures compared to private practices or community health centers. Furthermore, the reliance on EHR data for defining deprescribing (tapering vs. discontinuation) may not capture patient-initiated cessation that wasn’t recorded by the physician. However, the use of a randomized design and the focus on high-risk medication classes provide a strong foundation for broader implementation.
Conclusion and Summary
The findings of this randomized clinical trial support the integration of behavioral science-designed tools within EHR systems to improve the quality of care for older adults. By significantly increasing the rates of deprescribing for potentially inappropriate medications, these interventions offer a scalable solution to the epidemic of polypharmacy.As health systems continue to seek ways to improve patient safety and reduce unnecessary drug costs, the ‘nudge’ toward deprescribing may become a standard component of clinical informatics. The study underscores that while prescribing is an act of commission, deprescribing is an act of clinical courage and precision—one that can be effectively supported by well-designed technology.
Funding and ClinicalTrials.gov
This study was supported by grants from the National Institute on Aging (NIA) and other academic research funds. Trial registration: ClinicalTrials.gov Identifier: NCT05538065.
References
1. Lauffenburger JC, Sung M, Glynn RJ, et al. Electronic Health Record Intervention and Deprescribing for Older Adults: A Randomized Clinical Trial. JAMA. 2026;335(4):e2526967. doi:10.1001/jama.2025.26967.
2. By the 2023 American Geriatrics Society Beers Criteria Update Expert Panel. American Geriatrics Society 2023 updated AGS Beers Criteria for potentially inappropriate medication use in older adults. J Am Geriatr Soc. 2023;71(7):2052-2081.
3. Thaler RH, Sunstein CR. Nudge: Improving Decisions About Health, Wealth, and Happiness. Yale University Press; 2008.