ABSTRACT
BACKGROUND: Misuse of prescription opioids is a well-established contributor to the US opioid epidemic. The primary objective of this study was to identify which level of care delivery (ie patient, prescriber, or hospital) produced the most unwarranted variation in opioid prescribing after common surgical procedures. STUDY DESIGN: Electronic health record data from a large multihospital healthcare system were used in conjunction with random-effect models to examine variation in opioid prescribing practices after similar inpatient and outpatient surgical procedures between October 2019 and September 2021. Unwarranted variation was conceptualized as variation resulting from prescriber behavior unsupported by evidence. Covariates identified as drivers of warranted variation included characteristics known to influence pain levels or patient safety. All other model variables, including prescriber specialty and patient race, ethnicity, and insurance status were characterized as potential drivers of unwarranted variation. RESULTS: Among 25,188 procedures with an opioid prescription at hospital discharge, 53.5% exceeded guideline recommendations, corresponding to 13,228 patients receiving the equivalent of >140,000 excess 5 mg oxycodone tablets after surgical procedures. Prescribing variation was primarily driven by prescriber-level factors, with approximately half of the total variation in morphine milligram equivalents prescribed observed at the prescriber level and not explained by any measured variables. Unwarranted covariates associated with higher prescribed opioid quantity included non-Hispanic Black race, Medicare insurance, smoking history, later hospital discharge times, and prescription by a surgeon rather than a hospitalist or primary care provider. CONCLUSIONS: Given the large proportion of unexplained variation observed at the provider level, targeting prescribers through education and training may be an effective strategy for reducing postoperative opioid prescribing.
Subject(s)
Analgesics, Opioid , Pain, Postoperative , Practice Patterns, Physicians' , Humans , Analgesics, Opioid/therapeutic use , Pain, Postoperative/drug therapy , Practice Patterns, Physicians'/statistics & numerical data , Male , Female , Middle Aged , Adult , Aged , United States , Multi-Institutional Systems , Drug Prescriptions/statistics & numerical data , Surgical Procedures, Operative/statistics & numerical dataABSTRACT
Importance: Excess opioid prescribing after surgery can result in prolonged use and diversion. Email feedback based on social norms may reduce the number of pills prescribed. Objective: To assess the effectiveness of 2 social norm-based interventions on reducing guideline-discordant opioid prescribing after surgery. Design, Setting, and Participants: This cluster randomized clinical trial conducted at a large health care delivery system in northern California between October 2021 and October 2022 included general, obstetric/gynecologic, and orthopedic surgeons with patients aged 18 years or older discharged to home with an oral opioid prescription. Interventions: In 19 hospitals, 3 surgical specialties (general, orthopedic, and obstetric/gynecologic) were randomly assigned to a control group or 1 of 2 interventions. The guidelines intervention provided email feedback to surgeons on opioid prescribing relative to institutionally endorsed guidelines; the peer comparison intervention provided email feedback on opioid prescribing relative to that of peer surgeons. Emails were sent to surgeons with at least 2 guideline-discordant prescriptions in the previous month. The control group had no intervention. Main Outcome and Measures: The probability that a discharged patient was prescribed a quantity of opioids above the guideline for the respective procedure during the 12 intervention months. Results: There were 38â¯235 patients discharged from 640 surgeons during the 12-month intervention period. Control-group surgeons prescribed above guidelines 36.8% of the time during the intervention period compared with 27.5% and 25.4% among surgeons in the peer comparison and guidelines arms, respectively. In adjusted models, the peer comparison intervention reduced guideline-discordant prescribing by 5.8 percentage points (95% CI, -10.5 to -1.1; P = .03) and the guidelines intervention reduced it by 4.7 percentage points (95% CI, -9.4 to -0.1; P = .05). Effects were driven by surgeons who performed more surgeries and had more guideline-discordant prescribing at baseline. There was no significant difference between interventions. Conclusions and Relevance: In this cluster randomized clinical trial, email feedback based on either guidelines or peer comparison reduced opioid prescribing after surgery. Guideline-based feedback was as effective as peer comparison-based feedback. These interventions are simple, low-cost, and scalable, and may reduce downstream opioid misuse. Trial Registration: ClinicalTrials.gov NCT05070338.
Subject(s)
Analgesics, Opioid , Opioid-Related Disorders , Humans , Female , Analgesics, Opioid/therapeutic use , Feedback , Practice Patterns, Physicians' , Opioid-Related Disorders/drug therapy , PrescriptionsABSTRACT
BACKGROUND: Effective strategies are needed to curtail overuse that may lead to harm. OBJECTIVE: To evaluate the effects of clinician decision support redirecting attention to harms and engaging social and reputational concerns on overuse in older primary care patients. DESIGN: 18-month, single-blind, pragmatic, cluster randomized trial, constrained randomization. (ClinicalTrials.gov: NCT04289753). SETTING: 60 primary care internal medicine, family medicine and geriatrics practices within a health system from 1 September 2020 to 28 February 2022. PARTICIPANTS: 371 primary care clinicians and their older adult patients from participating practices. INTERVENTION: Behavioral science-informed, point-of-care, clinical decision support tools plus brief case-based education addressing the 3 primary clinical outcomes (187 clinicians from 30 clinics) were compared with brief case-based education alone (187 clinicians from 30 clinics). Decision support was designed to increase salience of potential harms, convey social norms, and promote accountability. MEASUREMENTS: Prostate-specific antigen (PSA) testing in men aged 76 years and older without previous prostate cancer, urine testing for nonspecific reasons in women aged 65 years and older, and overtreatment of diabetes with hypoglycemic agents in patients aged 75 years and older and hemoglobin A1c (HbA1c) less than 7%. RESULTS: At randomization, mean clinic annual PSA testing, unspecified urine testing, and diabetes overtreatment rates were 24.9, 23.9, and 16.8 per 100 patients, respectively. After 18 months of intervention, the intervention group had lower adjusted difference-in-differences in annual rates of PSA testing (-8.7 [95% CI, -10.2 to -7.1]), unspecified urine testing (-5.5 [CI, -7.0 to -3.6]), and diabetes overtreatment (-1.4 [CI, -2.9 to -0.03]) compared with education only. Safety measures did not show increased emergency care related to urinary tract infections or hyperglycemia. An HbA1c greater than 9.0% was more common with the intervention among previously overtreated diabetes patients (adjusted difference-in-differences, 0.47 per 100 patients [95% CI, 0.04 to 1.20]). LIMITATION: A single health system limits generalizability; electronic health data limit ability to differentiate between overtesting and underdocumentation. CONCLUSION: Decision support designed to increase clinicians' attention to possible harms, social norms, and reputational concerns reduced unspecified testing compared with offering traditional case-based education alone. Small decreases in diabetes overtreatment may also result in higher rates of uncontrolled diabetes. PRIMARY FUNDING SOURCE: National Institute on Aging.
Subject(s)
Diabetes Mellitus , Prostatic Neoplasms , Male , Humans , Aged , Prostate-Specific Antigen , Single-Blind Method , Hypoglycemic AgentsABSTRACT
We examined 3,046,538 acute respiratory infection (ARI) encounters with 6,103 national telehealth physicians from January 2019 to October 2021. The antibiotic prescribing rates were 44% for all ARIs; 46% were antibiotic appropriate; 65% were potentially appropriate; 19% resulted from inappropriate diagnoses; and 10% were related to coronavirus disease 2019 (COVID-19) diagnosis.
Subject(s)
Anti-Bacterial Agents , COVID-19 , Practice Patterns, Physicians' , Respiratory Tract Infections , Telemedicine , Humans , Anti-Bacterial Agents/therapeutic use , COVID-19/epidemiology , Respiratory Tract Infections/drug therapy , Practice Patterns, Physicians'/statistics & numerical data , Female , Male , Middle Aged , Aged , Adult , SARS-CoV-2 , Inappropriate Prescribing/statistics & numerical data , Young Adult , United States/epidemiology , Acute Disease , Drug Prescriptions/statistics & numerical data , AdolescentABSTRACT
Objectives. To examine whether the addition of telehealth data to existing surveillance infrastructure can improve forecasts of cases and mortality. Methods. In this observational study, we compared accuracy of 14-day forecasts using real-time data available to the National Syndromic Surveillance Program (standard forecasts) to forecasts that also included telehealth information (telehealth forecasts). The study was performed in a national telehealth service provider in 2020 serving 50 US states and the District of Columbia. Results. Among 10.5 million telemedicine encounters, 169 672 probable COVID-19 cases were diagnosed by 5050 clinicians, with a rate between 0.79 and 47.8 probable cases per 100 000 encounters per day (mean = 8.37; SD = 10.75). Publicly reported case counts ranged from 0.5 to 237 916 (mean: 53 913; SD = 47 466) and 0 to 2328 deaths (mean = 1035; SD = 550) per day. Telehealth-based forecasts improved 14-day case forecasting accuracy by 1.8 percentage points to 30.9% (P = .06) and mortality forecasting by 6.4 percentage points to 26.9% (P < .048). Conclusions. Modest improvements in forecasting can be gained from adding telehealth data to syndromic surveillance infrastructure. (Am J Public Health. 2024;114(2):218-225. https://doi.org/10.2105/AJPH.2023.307499).
Subject(s)
COVID-19 , Telemedicine , Humans , COVID-19/epidemiology , Pandemics , Telemedicine/methods , District of Columbia , ForecastingABSTRACT
Objective: Patient-reported outcome measures (PROMs) are critical to drive patient-centered care and to understanding patients' perspectives on their health status, quality of life, and the overall effectiveness of the care they receive. PROMs are increasingly being used in clinical and research settings, but the mechanisms to aggregate data from different systems can be cumbersome. Materials and methods: As part of an FDA Real-World Evidence demonstration project, we enriched routine care clinical data from our Cerner electronic health record (EHR) with PROMs collected using REDCap. We used SSIS, sFTP, and the REDCap Application Programming Interface to aggregate both data sources into the Cerner HealtheIntent Population Health Platform. Results: We successfully built dashboards, reports, and datasets containing both REDCap and EHR data collected prospectively. Discussion: This technically straightforward approach using commonly available clinical and research tools can be readily adopted and adapted by others to better integrate PROMs with clinical data sources.
ABSTRACT
Importance: Interventions that improve clinician performance through feedback should not contribute to job dissatisfaction or staff turnover. Measurement of job satisfaction may help identify interventions that lead to this undesirable consequence. Objective: To evaluate whether mean job satisfaction was less than the margin of clinical significance among clinicians who received social norm feedback (peer comparison) compared with clinicians who did not. Design, Setting, and Participants: This secondary, preregistered, noninferiority analysis of a cluster randomized trial compared 3 interventions to reduce inappropriate antibiotic prescribing in a 2 × 2 × 2 factorial design from November 1, 2011, to April 1, 2014. A total of 248 clinicians were enrolled from 47 clinics. The sample size for this analysis was determined by the number of nonmissing job satisfaction scores from the original enrolled sample, which was 201 clinicians from 43 clinics. Data analysis was performed from October 12 to April 13, 2022. Interventions: Feedback comparing individual clinician performance to top-performing peers, delivered in monthly emails (peer comparison). Main Outcomes and Measures: The primary outcome was a response to the following statement: "Overall, I am satisfied with my current job." Responses ranged from 1 (strongly disagree) to 5 (strongly agree). Results: A total of 201 clinicians (response rate, 81%) from 43 of the 47 clinics (91%) provided a survey response about job satisfaction. Clinicians were primarily female (n = 129 [64%]) and board certified in internal medicine (n = 126 [63%]), with a mean (SD) age of 48 (10) years. The clinic-clustered difference in mean job satisfaction was greater than -0.32 (ß = 0.11; 95% CI, -0.19 to 0.42; P = .46). Therefore, the preregistered null hypothesis that peer comparison is inferior by resulting in at least a 1-point decrease in job satisfaction by 1 in 3 clinicians was rejected. The secondary null hypothesis that job satisfaction was similar among clinicians randomized to social norm feedback was not able to be rejected. The effect size did not change when controlling for other trial interventions (t = 0.08; P = .94), and no interaction effects were observed. Conclusions and Relevance: In this secondary analysis of a randomized clinical trial, peer comparison did not lead to lower job satisfaction. Features that may have protected against dissatisfaction include clinicians' agency over the performance measure, privacy of individual performance, and allowing all clinicians to achieve top performance. Trial Registration: ClinicalTrials.gov Identifiers: NCT05575115 and NCT01454947.
Subject(s)
Emotions , Job Satisfaction , Humans , Female , Middle Aged , Surveys and Questionnaires , Feedback , Anti-Bacterial Agents/therapeutic useABSTRACT
Common data models solve many challenges of standardizing electronic health record (EHR) data but are unable to semantically integrate all of the resources needed for deep phenotyping. Open Biological and Biomedical Ontology (OBO) Foundry ontologies provide computable representations of biological knowledge and enable the integration of heterogeneous data. However, mapping EHR data to OBO ontologies requires significant manual curation and domain expertise. We introduce OMOP2OBO, an algorithm for mapping Observational Medical Outcomes Partnership (OMOP) vocabularies to OBO ontologies. Using OMOP2OBO, we produced mappings for 92,367 conditions, 8611 drug ingredients, and 10,673 measurement results, which covered 68-99% of concepts used in clinical practice when examined across 24 hospitals. When used to phenotype rare disease patients, the mappings helped systematically identify undiagnosed patients who might benefit from genetic testing. By aligning OMOP vocabularies to OBO ontologies our algorithm presents new opportunities to advance EHR-based deep phenotyping.
ABSTRACT
BACKGROUND AND OBJECTIVE: Social and Environmental Determinants of Health (SEDoH) are of increasing interest to researchers in personal and public health. Collecting SEDoH and associating them with patient medical record can be challenging, especially for environmental variables. We announce here the release of SEnDAE, the Social and Environmental Determinants Address Enhancement toolkit, and open-source resource for ingesting a range of environmental variables and measurements from a variety of sources and associated them with arbitrary addresses. METHODS: SEnDAE includes optional components for geocoding addresses, in case an organization does not have independent capabilities in that area, and recipes for extending the OMOP CDM and the ontology of an i2b2 instance to display and compute over the SEnDAE variables within i2b2. RESULTS: On a set of 5000 synthetic addresses, SEnDAE was able to geocode 83%. SEnDAE geocodes addresses to the same Census tract as ESRI 98.1% of the time. CONCLUSION: Development of SEnDAE is ongoing, but we hope that teams will find it useful to increase their usage of environmental variables and increase the field's general understanding of these important determinants of health.
Subject(s)
Medical Records , Public Health , HumansABSTRACT
BACKGROUND: Unnecessary testing and treatment of common conditions in older adults can lead to significant morbidity and mortality. The primary objective of this study was to develop and pilot test a set of clinical decision support (CDS) alerts informed by social psychology to address overuse in three areas related to ambulatory care of older adults. METHODS: We developed three electronic health record (EHR) CDS alerts to address overuse and pilot tested them from January 17, 2019 to July 17, 2019. We enrolled 14 primary care physicians from three practices within a large health system who cared for adults aged 65 years and older. Three measures of overuse applied to patients meeting the following criteria: ordering of prostate-specific antigen (PSA) for prostate cancer screening in adult men aged 76 years and older, ordering of urinalysis or urine cultures (UA or UC) for non-specific reasons to identify bacteriuria in women aged 65 years and older, and overtreatment of diabetes with insulin or oral hypoglycemic medications in adults aged at 75 years and older (DM). Clinicians received CDS alerts when criteria for any of the three overuse measures were met. We then surveyed clinicians to evaluate their experience with the CDS alerts. RESULTS: The number of clinical encounters that triggered CDS alerts was 19 for PSA, 48 for UA/UC and 128 for DM. For PSA encounters, 4 (21%) orders were not performed after the alert. In the UA/UC encounters 29 (60%) orders were not performed after the alert. For the DM encounters, 21 (34%) had diabetes therapy reduced following the alert. Survey respondents indicated that the alerts were clinically accurate and sometimes led them to change their clinical action. CONCLUSIONS: These CDS alerts were feasible to implement and may minimize unnecessary testing and treatment of common conditions in older adults.
Subject(s)
Electronic Health Records , Prostatic Neoplasms , Male , Humans , Aged , Prostate-Specific Antigen , Early Detection of Cancer , Primary Health CareABSTRACT
Variation in availability, format, and standardization of patient attributes across health care organizations impacts patient-matching performance. We report on the changing nature of patient-matching features available from 2010-2020 across diverse care settings. We asked 38 health care provider organizations about their current patient attribute data-collection practices. All sites collected name, date of birth (DOB), address, and phone number. Name, DOB, current address, social security number (SSN), sex, and phone number were most commonly used for cross-provider patient matching. Electronic health record queries for a subset of 20 participating sites revealed that DOB, first name, last name, city, and postal codes were highly available (>90%) across health care organizations and time. SSN declined slightly in the last years of the study period. Birth sex, gender identity, language, country full name, country abbreviation, health insurance number, ethnicity, cell phone number, email address, and weight increased over 50% from 2010 to 2020. Understanding the wide variation in available patient attributes across care settings in the United States can guide selection and standardization efforts for improved patient matching in the United States.
ABSTRACT
Introduction: Integrating social and environmental determinants of health (SEDoH) into enterprise-wide clinical workflows and decision-making is one of the most important and challenging aspects of improving health equity. We engaged domain experts to develop a SEDoH informatics maturity model (SIMM) to help guide organizations to address technical, operational, and policy gaps. Methods: We established a core expert group consisting of developers, informaticists, and subject matter experts to identify different SIMM domains and define maturity levels. The candidate model (v0.9) was evaluated by 15 informaticists at a Center for Data to Health community meeting. After incorporating feedback, a second evaluation round for v1.0 collected feedback and self-assessments from 35 respondents from the National COVID Cohort Collaborative, the Center for Leading Innovation and Collaboration's Informatics Enterprise Committee, and a publicly available online self-assessment tool. Results: We developed a SIMM comprising seven maturity levels across five domains: data collection policies, data collection methods and technologies, technology platforms for analysis and visualization, analytics capacity, and operational and strategic impact. The evaluation demonstrated relatively high maturity in analytics and technological capacity, but more moderate maturity in operational and strategic impact among academic medical centers. Changes made to the tool in between rounds improved its ability to discriminate between intermediate maturity levels. Conclusion: The SIMM can help organizations identify current gaps and next steps in improving SEDoH informatics. Improving the collection and use of SEDoH data is one important component of addressing health inequities.
ABSTRACT
BACKGROUND: Requiring accountable justifications-visible, clinician-recorded explanations for not following a clinical decision support (CDS) alert-has been used to steer clinicians away from potentially guideline-discordant decisions. Understanding themes from justifications across clinical content areas may reveal how clinicians rationalize decisions and could help inform CDS alerts. METHODS: We conducted a qualitative evaluation of the free-text justifications entered by primary care physicians from three pilot interventions designed to reduce opioid prescribing and, in older adults, high-risk polypharmacy and overtesting. Clinicians encountered alerts when triggering conditions were met within the chart. Clinicians were asked to change their course of action or enter a justification for the action that would be displayed in the chart. We extracted all justifications and grouped justifications with common themes. Two authors independently coded each justification and resolved differences via discussion. Three physicians used a modified Delphi technique to rate the clinical appropriateness of the justifications. RESULTS: There were 560 justifications from 50 unique clinicians. We grouped these into three main themes used to justify an action: (1) report of a particular diagnosis or symptom (e.g., for "anxiety" or "acute pain"); (2) provision of further contextual details about the clinical case (e.g., tried and failed alternatives, short-term supply, or chronic medication); and (3) noting communication between clinician and patient (e.g., "risks and benefits discussed"). Most accountable justifications (65%) were of uncertain clinical appropriateness. CONCLUSION: Most justifications clinicians entered across three separate clinical content areas fit within a small number of themes, and these common rationales may aid in the design of effective accountable justification interventions. Justifications varied in terms of level of clinical detail. On their own, most justifications did not clearly represent appropriate clinical decision making.
Subject(s)
Decision Support Systems, Clinical , Physicians , Aged , Analgesics, Opioid , Humans , Practice Patterns, Physicians' , Social ResponsibilityABSTRACT
Simulation is a mainstay of comparative- and cost-effectiveness research when empirical data are not available. The Synthea platform, originally designed for generating realistically coded longitudinal health records for software testing, implements data generation models specified in publicly contributed modules representing patients' life cycle and disease and treatment progression. We test the hypothesis that Synthea can be used for simulation studies that draw parameters from observational studies and randomized trials. We benchmarked the results and assessed the effort required to create a Synthea module that replicates a recently published cost-effectiveness simulation comparing levofloxacin prophylaxis to usual care for leukemia. A module was iteratively developed using published parameters from the original study; we replicated the initial conditions and simulation endpoints of demographics, health events, costs, and mortality. We compare Synthea's Generic Module Framework to platforms designed for simulation and show that Synthea can be used, with modifications, for some types of simulation studies.
ABSTRACT
Including social determinants of health (SDoH) data in health outcomes research is essential for studying the sources of healthcare disparities and developing strategies to mitigate stressors. In this report, we describe a pragmatic design and approach to explore the encoding needs for transmitting SDoH screening tool responses from a large safety-net hospital into the National Covid Cohort Collaborative (N3C) OMOP dataset. We provide a stepwise account of designing data mapping and ingestion for patient-level SDoH and summarize the results of screening. Our approach demonstrates that sharing of these important data - typically stored as non-standard, EHR vendor specific codes - is feasible. As SDoH screening gains broader use nationally, the approach described in this paper could be used for other screening instruments and improve the interoperability of these important data.
ABSTRACT
BACKGROUND: The CDC estimates that over 40% of Urgent Care visits are for acute respiratory infections (ARI), more than half involving inappropriate antibiotic prescriptions. Previous randomized trials in primary care clinics resulted in reductions in inappropriate antibiotic prescribing, but antibiotic stewardship interventions in telehealth have not been systematically assessed. To better understand how best to decrease inappropriate antibiotic prescribing for ARIs in telehealth, we are conducting a large randomized quality improvement trial testing both patient- and physician-facing feedback and behavioral nudges embedded in the electronic health record. METHODS: Teladoc® clinicians are assigned to one of 9 arms in a 3 × 3 randomized trial. Each clinician is assigned to one of 3 Commitment groups (Public, Private, Control) and one of 3 Performance Feedback groups (Benchmark Peer Comparison, Trending, Control). After randomly selecting â of states and associated clinicians required for patient-facing components of the Public Commitment intervention, remaining clinicians are randomized to the Control and Private Commitment arms. Clinicians are randomized to the Performance Feedback conditions. The primary outcome is change from baseline in antibiotic prescribing rate for qualifying ARI visits. Secondary outcomes include changes in inappropriate prescribing and revisit rates. Secondary analyses include investigation of heterogeneity of treatment effects. With 1530 clinicians and an intra-clinician correlation in antibiotic prescribing rate of 0.5, we have >80% power to detect 1-7% absolute differences in antibiotic prescribing among groups. DISCUSSION: Findings from this trial may help inform telehealth stewardship strategies, determine whether significant differences exist between Commitment and Feedback interventions, and provide guidance for clinicians and patients to encourage safe and effective antibiotic use. CLINICALTRIALS: gov: NCT05138874.
Subject(s)
Respiratory Tract Infections , Telemedicine , Anti-Bacterial Agents , Electronic Health Records , Humans , Inappropriate Prescribing , Practice Patterns, Physicians' , Randomized Controlled Trials as TopicABSTRACT
While much is known about governance models for research informatics programs in academic medical centers and similarly situated cancer centers, community and public health systems have been less well-characterized. As part of implementing an enterprise research governance framework, leaders in the Los Angeles County Department of Health Services established a research informatics program, including research data warehousing. The strategy is focused on high-priority, patient-centered research that leverages the investment in health IT and an efficient, sustained contribution from 2 affiliated Clinical Translational Sciences Institutes. This case study describes the foundational governance framework and policies that were developed. We share the results of several years of planning, implementation, and operations of an academically funded research informatics service core embedded in a large, multicenter county health system. We include herein a Supplementary Appendix of governance documents that may serve as pragmatic models for similar initiatives.
Subject(s)
Delivery of Health Care , Informatics , Academic Medical Centers , Data Warehousing , HumansABSTRACT
BACKGROUND: Inappropriate polypharmacy, prevalent among older patients, is associated with substantial harms. OBJECTIVE: To develop measures of high-risk polypharmacy and pilot test novel electronic health record (EHR)-based nudges grounded in behavioral science to promote deprescribing. DESIGN: We developed and validated seven measures, then conducted a three-arm pilot from February to May 2019. PARTICIPANTS: Validation used data from 78,880 patients from a single large health system. Six physicians were pre-pilot test environment users. Sixty-nine physicians participated in the pilot. MAIN MEASURES: Rate of high-risk polypharmacy among patients aged 65 years or older. High-risk polypharmacy was defined as being prescribed ≥5 medications and satisfying ≥1 of the following high-risk criteria: drugs that increase fall risk among patients with fall history; drug-drug interactions that increase fall risk; thiazolidinedione, NSAID, or non-dihydropyridine calcium channel blocker in heart failure; and glyburide, glimepiride, or NSAID in chronic kidney disease. INTERVENTIONS: Physicians received EHR alerts when renewing or prescribing certain high-risk medications when criteria were met. One practice received a "commitment nudge" that offered a chance to commit to addressing high-risk polypharmacy at the next visit. One practice received a "justification nudge" that asked for a reason when high-risk polypharmacy was present. One practice received both. KEY RESULTS: Among 55,107 patients 65 and older prescribed 5 or more medications, 6256 (7.9%) had one or more high-risk criteria. During the pilot, the mean (SD) number of nudges per physician per week was 1.7 (0.4) for commitment, 0.8 (0.5) for justification, and 1.9 (0.5) for both interventions. Physicians requested to be reminded to address high-risk polypharmacy for 236/833 (28.3%) of the commitment nudges and acknowledged 441 of 460 (95.9%) of justification nudges, providing a text response for 187 (40.7%). CONCLUSIONS: EHR-based measures and nudges addressing high-risk polypharmacy were feasible to develop and implement, and warrant further testing.