EHR Behavioral Prompts Increase Medication Deprescribing in Older Adults
-
Click here to listen to this article - Share via
- Primary care physicians receiving sequential EHR nudges were 40% more likely to discontinue or taper potentially inappropriate medications in older adults.
- The trial targeted high-risk drug classes, including benzodiazepines, nonbenzodiazepine sedative hypnotics, and strongly anticholinergic medications.
- Precommitment prompts encouraged clinicians to initiate deprescribing discussions during an initial visit and follow through at a subsequent visit.
- Deprescribing occurred without an increase in serious adverse events across the study population.
- Behavioral science nudges provide a practical method to reduce medication burden without adding time to routine clinical appointments.
Table of Contents
- Behavioral Nudges to Reduce Potentially Inappropriate Medications
- How Precommitment Prompts Overcome Clinical Inertia
- Safety Profile and Patient Outcomes
- Closing the Research Gap in Primary Care
- Closing Thoughts
Behavioral Nudges to Reduce Potentially Inappropriate Medications
Potentially inappropriate medications remain widely prescribed to adults aged 65 and older despite established safety risks. Long-term use of benzodiazepines, nonbenzodiazepine sleep aids, and strongly anticholinergic drugs increases the risk of cognitive decline, physical falls, and hospital admissions. Clinicians often face barriers to stopping these drugs, including brief visit times, patient resistance, and clinical inertia.
To evaluate whether behavioral science tools embedded in health IT could change prescribing habits, researchers conducted a cluster-randomized clinical trial across primary care clinics [1]. The study tested whether behaviorally designed EHR prompts deprescribe sedatives older adults take long-term more effectively than standard care.
Quick Take & Core Data
- Precommitment EHR prompts increased deprescribing rates by 40% compared with usual care (36.8% vs. 26.8%; relative risk 1.40; 10.4% absolute increase) [1].
- Boostering prompts with optional in-basket reminders increased deprescribing by 26% compared with usual care (34.3% vs. 26.8%; relative risk 1.26; 6.5% absolute increase) [1].
- 1,146 patients aged 65 or older (mean age 73.6 years; 69.7% female) managed by 201 primary care physicians over a mean follow-up of 289.9 days [1].
- Zero serious adverse events were reported through formal safety reporting mechanisms [1].
How Precommitment Prompts Overcome Clinical Inertia
Traditional electronic alerts often fail because they appear as passive pop-ups that clinicians quickly dismiss. This study tested two distinct nudge strategies grounded in behavioral economics.
The first strategy used precommitment prompts. During a patient visit, the EHR displayed a message asking the physician to start a deprescribing conversation with the patient. If the physician agreed, the EHR delivered a second reminder at the next visit to complete the tapering or discontinuation plan [1].
The second strategy, known as boostering, provided an initial notification during the visit and allowed physicians to opt into receiving a follow-up reminder in their EHR inbox four weeks later. In usual care, physicians received no automated prompts regarding medication discontinuation.
Dae Hyun Kim, MD, MPH, ScD, a geriatrics researcher at Hebrew SeniorLife’s Marcus Institute for Aging Research and co-author of the study, noted in a statement provided by the institution that many older adults remain on medications long past the point where risks outweigh benefits.
Kim stated that thoughtfully designed electronic health record tools can support better conversations between clinicians and patients, making deprescribing a more routine part of primary care without adding burden to visits.
Safety Profile and Patient Outcomes
Over a mean follow-up period of nearly ten months, 32.5% of all enrolled patients had at least one targeted medication discontinued or tapered [1]. Deprescribing occurred in 36.8% of patients in the precommitment group, 34.3% in the boostering group, and 26.8% in usual care [1].
The precommitment nudge achieved statistical significance after adjusting for baseline characteristics and clinic clustering [1].
Safety monitoring showed no difference in harm between the intervention arms and usual care. No serious adverse events occurred via formal reporting channels [1]. Mortality rates identified through manual chart review were 1.4% in the precommitment group, 3.9% in the boostering group, and 1.8% in usual care [1].
Closing the Research Gap in Primary Care
Prior research established that electronic decision support can increase the uptake of recommended medications. However, evidence was limited regarding whether health record prompts could successfully guide physicians to stop established, non-recommended treatments.
Standard alert systems frequently trigger alert fatigue, leading clinicians to ignore automated warnings.
This trial fills that gap by demonstrating that two-stage behavioral nudges, which match clinical decision-making steps, successfully alter physician practice without interrupting workflow [1].
A primary limitation of the study is its execution within a single academic health system in Massachusetts, which may limit generalizability to community health centers or different EHR platforms [1].
Additionally, outcomes relied on EHR documentation, which might miss unrecorded medication changes or informal patient-led dose reductions [1].
Closing Thoughts
Integrating behavioral science principles into health record systems provides a proven method to reduce high-risk medication use in older populations.
Sequential precommitment prompts give primary care physicians a practical mechanism to initiate and follow through on complex drug discontinuations.
As health systems evaluate practical approaches to improve geriatric care quality, software-based nudges provide an effective path to match daily practice with safety guidelines.
References
[1] Lauffenburger, J. C., Sung, M., Glynn, R. J., Keller, P. A., Robertson, T., Kim, D. H., Bhatkhande, G., Jungo, K. T., Haff, N., Hanken, K. E., Isaac, T., & Choudhry, N. K. (2026). Electronic Health Record Intervention and Deprescribing for Older Adults: A Randomized Clinical Trial. *JAMA*, 335(12), 1060–1069. https://doi.org/10.1001/jama.2025.26967
