Reducing High Risk Polypharmacy Using Behavioral Economics through Electronic Health Records

High-risk polypharmacy is common among older adults in the United States and is associated with harms such as adverse drug reactions, falls, and higher costs of care. Individuals with cognitive impairment or dementia may be at particularly at risk for adverse outcomes from polypharmacy. To date, efforts to reduce the rate of high-risk polypharmacy have had limited uptake due to their expense per clinician and their modest effectiveness in helping clinicians stop potentially inappropriate medications. These prior efforts have relied on rational clinician behavior models. However, like people in general, clinicians are susceptible to irrational biases and cognitive shortcuts that might perpetuate inappropriate high-risk polypharmacy. Therefore, behavioral economic nudges, which do not assume rational decision making, might be well-positioned to reduce such high-risk polypharmacy. Moreover, such nudges can be embedded in electronic health records (EHRs), enhancing their scalability at a low cost per clinician—as long as they are acceptable to clinicians who worry about EHRs’ effects on their workflows. The overall objectives of the project are to test EHR-based behavioral economic nudges that may reduce the prevalence and degree of high-risk polypharmacy among older adult patients and to describe their impact on clinicians’ experiences and workflows. The specific aims are (1) to evaluate the effects of an EHR-based commitment nudge, a justification nudge, and the combination of both nudges on a composite measure of high-risk polypharmacy in all older adults and the subpopulation with dementia or mild cognitive impairment via a pragmatic randomized controlled trial. This will use cluster randomization in which primary care clinics from two large health systems are randomized to receive 0, 1, or 2 nudges using a factorial design. The nudges will run for 18 months, followed by 12 months of observation to assess persistence of effects; and (2) to assess qualitatively and quantitatively clinician experiences with the EHR-based nudges, including their acceptability and effects on workflow. At the conclusion of the intervention period, semi-structured interviews with approximately 40 clinicians will be used to describe their experiences with the nudges in detail, and a clinician survey will be conducted to examine the relationships between clinicians’ perceptions of the nudges, their actual use of the EHR tools, and the relationships of these factors to actual changes in high-risk polypharmacy. The study will yield EHR-based nudges that can be implemented across the country, detailed descriptions of how these nudges affect clinicians’ workflows (to spur dissemination), and scientific papers documenting the nudges’ effectiveness in reducing high-risk polypharmacy among older adults.