Publications

OBJECTIVE: To investigate whether eligibility for Veterans Health Administration (VA)-purchased community care, which expanded Veterans' access to care outside VA, was associated with increased polypharmacy or potentially inappropriate medication use among older adult Veterans.

STUDY SETTING AND DESIGN: Regression discontinuity design, leveraging the distance threshold for community care eligibility (residing > 40 miles from the nearest VA facility with ≥ 1 or more full-time primary care physician), to examine the effects of community care eligibility on polypharmacy and potentially inappropriate medication use among Veterans aged ≥ 65 years.

DATA SOURCES AND ANALYTIC SAMPLE: VA pharmacy data for all prescriptions filled at VA facilities, VA Program Integrity Tool files for prescriptions paid by VA and filled in community pharmacies, and Medicare Part D data. Analyses included annual cross-sectional samples of Veterans 36-39 miles or 41-44 miles from their nearest VA facility during FY 2016-2019.

PRINCIPAL FINDINGS: The sample included 399,250 Veteran-year observations, of which 226,157 (56.6%) were 36-39 miles and 173,093 (43.4%) were 41-44 miles from the nearest eligible VA facility. Overall, we observed no discontinuities across the 40-mile threshold in the number of unique medications filled annually (-0.06 medications; 95% confidence interval [CI], -0.15 to 0.03). There were no discontinuities in proportions of Veterans filling ≥ 5 unique medications (-0.25 percentage points [pp]; 95% CI, -0.84 to 0.34), ≥ 10 medications (-0.55 pp.; 95% CI, -1.24 to 0.14), ≥ 1 medication on the Beers list (-0.02 pp.; 95% CI, -0.63 to 0.59), or ≥ 1 high-risk drug-drug interaction (-0.05 pp.; 95% CI, -0.17 to 0.07). Among Veterans with mental health conditions, exceeding the 40-mile threshold was associated with a higher likelihood of filling ≥ 10 unique medications annually (2.06 pp.; 95% CI, 0.42 to 3.70). We did not observe clinically or statistically significant discontinuities in other subgroups.

CONCLUSIONS: Overall, eligibility for VA-purchased community care was not associated with increased polypharmacy or potentially inappropriate medication use among older adult Veterans.

BACKGROUND: People with non-traumatic fractures typically qualify for antifracture medications to prevent recurrent fractures.

OBJECTIVE: To quantify the current osteoporosis care gap and determine factors that predict prescribing of antifracture medications following hospitalization for non-traumatic fracture.

DESIGN: Retrospective cohort study.

SETTING: Twenty hospitals from a large US health system.

PATIENTS: Patients aged 50 years and older presenting to the emergency department with a non-traumatic hip, pelvic, or vertebral fracture between 2017 and 2023.

MEASUREMENTS: In-hospital administration or outpatient prescription fill of antifracture medications within 180 days following hospital discharge. We estimated incident rate ratios (IRRs) using modified Poisson regression with robust standard errors to in order to determine factors associated with antifracture medication initiation.

RESULTS: The cohort included 9849 patients (median (IQR) age, 80 [70-87] years; 68% female) of which 9008 (91.4%) were antifracture medication naïve prior to hospital presentation. Of medication-naïve patients, 510 (5.7%) were prescribed antifracture medications, with 279 (3.1%) initiating treatment as an inpatient and 119 (1.3%) in the 180 days following discharge. Medication classes included bisphosphonate (N = 431, 85%), RANK ligand inhibitors and selective estrogen receptor modulators (N = 56, 11%), and anabolic therapies (N = 21, 4%). Female sex (IRR 1.93; 95% CI 1.41-2.64) and a pre-existing diagnosis of osteoporosis (IRR 1.76, 95% CI 1.47-2.11) were positively associated with medication receipt.

LIMITATIONS: Inability to measure reasons for treatment non-initiation.

CONCLUSIONS: This study found very few adults with a non-traumatic fracture received medical treatment for secondary osteoporosis fracture prevention. Expanded quality improvement efforts are needed to close the osteoporosis care gap.

Aggregated counts from electronic health records (EHRs) provide a rich source of real-world data for investigating critical medical conditions such as postpartum hemorrhage (PPH). We used the Evolve to Next-Gen Accrual to Clinical Trials (ENACT) network, a large, federated network of EHRs, to characterize national trends, risk factors, and comorbidities associated with PPH in a cohort of 705,120 women hospitalized for delivery from 2005 to 2022. During this period, there was a statistically significant increase in the incidence of PPH, consistent with previous studies. Asian women had the highest incidence of PPH (10.33%), followed by Black or African American (8.33%) and American Indian or Alaska Native (8.05%) women. In PPH deliveries, the top-ranked risk factor was severe eclampsia (18.51%), the top-ranked comorbidity was operative vaginal delivery (16.98%), and the commonest cause was uterine atony (83.06%). We demonstrated that a large federated EHR network such as ENACT can be used to generate population-level epidemiological findings, albeit with some limitations.

INTRODUCTION: A 2019 Medicare opioid safety policy limited initial opioid duration to 7 days and long-term daily dosage to 90 morphine milligram equivalence (MME). Because effective pain management may vary by rurality, this study examined whether the policy differentially affected opioid prescribing among rural versus urban beneficiaries with disabilities.

METHODS: Deidentified claims data was used to gather a rolling cohort of Medicare Advantage beneficiaries with disabilities ages 18 to 64 from 2016 to 2021. Comparative interrupted time series were used in 2025 to analyze rural-urban differences in the duration and dosages of opioid fills for new-to-opioid beneficiaries, the target of the 7-day limit (N = 526,019 person-months), and the number of high-dosage episodes for long-term opioid beneficiaries, the target of the 90-MME limit (N = 3,312,161 person-months).

RESULTS: Despite significant reductions in initial opioid duration for both groups, the 7-day limit effect on initial opioid duration weakened over time (trend change: 0.17 [98.75% CI 0.04 to 0.30]) in the rural relative to the urban group. By the end of the study period, the rural group experienced 3.8 percentage points more initial opioid fills exceeding the 7-day limit (95% CI 0.4 to 7.3) than the urban group. Furthermore, the total dosage prescribed in subsequent fills among rural beneficiaries was 27.97 MME (95% CI -52.19 to -3.75) less than urban beneficiaries by the end of the study period. The rural and urban groups responded similarly to the 90-MME limitation.

CONCLUSIONS: Rural relative to urban new-to-opioid beneficiaries with disabilities experienced a weakened effect on initial opioid duration after Medicare's 7-day limit, while receiving declining dosage in follow-up opioid fills, suggesting potential disparities in access to follow-up opioid prescriptions for rural versus urban beneficiaries with disabilities. Policymakers may consider more flexible opioid regulations and attention to alternative pain management for rural Medicare beneficiaries.

BACKGROUND: Despite limited evidence of efficacy and well-established risks, long-term benzodiazepine use remains common. In July 2016, the Oregon Medicaid program introduced a prior authorization policy limiting benzodiazepine prescriptions to 4 weeks.

OBJECTIVES: To evaluate the impact of Oregon's benzodiazepine policy on medication utilization and health outcomes.

RESEARCH DESIGN: Interrupted time series regressions from January 2015-December 2017 (18 mo pre vs. post-policy).

SUBJECTS: Rolling cohorts of adult benzodiazepine recipients with 6 months of continuous enrollment before and after each month of observation.

MEASURES: Benzodiazepine utilization included monthly measures of overall, prevalent long-term, and incident long-term medication use. Harms included a composite measure of emergency department or inpatient admissions for overdose, self-harm, acute withdrawal, or acute psychiatric events, and were compared with patients using similar medications that were not affected by the policy.

RESULTS: The study population included 69,143 benzodiazepine recipients; 68% were female, and the average age was 41 years or older. Following policy implementation, there was a significant segment decrease (-​​​29.82 recipients/1000 patients; 95% CI: -​​​33.75 to -​​​25.89) in incident long-term benzodiazepine use and a 1.17 recipients/1000 patients/month (95% CI: 0.82-1.52) increase in trend. Trends in prevalent long-term (-1.16 recipients/1000 patients/month: 95% CI: -1.32 to -0.99) benzodiazepine use also declined. Modest trend decreases in harm were observed among all cohorts.

CONCLUSIONS: A policy to limit benzodiazepine duration was associated with a substantial reduction in progression to longer-term use and modest declines in the trend for potential harms.

Medicaid is the largest insurer for patients with opioid use disorder (OUD) and finances a disproportionate share of medication treatment for OUD (MOUD). Within Medicaid, initiation and continuity of MOUD are associated with improved outcomes, but substantial inequities across populations and care settings remain. The Medicaid Outcomes Distributed Research Network (MODRN) has supported Medicaid agency efforts to improve MOUD receipt and continuity of care. In the context of MODRN-Examining Quality Improvement for Medicaid Programs, the Behavioral Health Treatment Engagement and Retention Learning Collaborative (BHTER) will implement and evaluate interventions focused on improving quality and reducing variation in process and clinical outcomes in Medicaid patients with OUD. BHTER's aims are to (1) estimate the effects of a quality improvement intervention on receipt and continuity of MOUD in intervention practices relative to comparison practices, (2) examine whether a quality improvement intervention narrows racial gaps in rates and continuity of MOUD in intervention relative to comparison practices, (3) examine provider perspectives on the quality improvement intervention, with a focus on practice and process changes that resulted in improvement, and factors that impede or facilitate delivery of high-quality MOUD care. In this protocol paper, we describe the rationale, methods, and impact of BHTER.

Public payers in the US cover more than half of individuals who receive treatment for opioid use disorder (OUD). Despite this, patient-reported measures that reflect treatment engagement and quality of life are largely missing from OUD quality improvement initiatives. This commentary describes the survey protocol for Project ELVIS (ELevating patient VoIceS to Improve OUD Treatment Quality), 1 of 3 projects of the Medicaid Outcomes Distributed Research Network's initiative Examining Quality Improvement in Medicaid Programs. Project ELVIS collects patient experiences with OUD treatment and tests the relationship between patient-reported and claims-based measures of treatment quality. As this commentary describes, Project ELVIS is innovative in its approach to grounding measure development in sustained, iterative engagement with individuals with lived experience and Medicaid policymakers and in employing a policy-driven regional survey sampling approach that prioritizes Medicaid decision-making relevance. Our project is also innovative in its planned linkages of self-reported patient experiences and outcomes with quality measures based on administrative data, positioning patient experience to explain-not just quantify-treatment quality in Medicaid programs. The goals of this project are to inform mechanisms underlying providers' high (and low) performance on claims-based measures, offering insight into why patients initiate, remain engaged in, or discontinue medication for OUD. In doing so, this work endeavors to support the development of patient-centered, process-based quality measures that complement existing and emerging utilization-based metrics.

BACKGROUND: Medicaid prescription cap policies persist in 12 states; however, it is unclear how they affect the quality of opioid use disorder (OUD) treatment.

OBJECTIVES: To examine the association between cap policies and the continuity in receipt of medication for OUD (MOUD) treatment in Medicaid.

METHODS: Using 2016-2021 T-MSIS Analytical Files data from 37 states, we identified nondual adult (18-64 y) Medicaid enrollees diagnosed with OUD who received MOUD. The exposure was the state's Medicaid prescription cap policy status. The main outcome was a binary variable for MOUD continuity, overall and by type, defined as no treatment gap exceeding 7 consecutive days during a 180-day observation period. We estimated risk ratios for MOUD continuity using modified Poisson regression models, adjusting for individual and state-level covariates to test associations.

RESULTS: Nearly half the sample was female, and the mean age was 37 years. MOUD continuity ranged from 40.7% to 44.3% in states without cap policies and 37.5%-39.6% in cap states. Compared with enrollees in noncap states, those in cap states had a 12% higher likelihood of experiencing at least 1 MOUD treatment gap [risk ratio (RR)=1.12; 95% CI: 1.12-1.13]. These findings were consistent for buprenorphine and methadone. Factors associated with overall MOUD discontinuity included younger age, male sex, living in nonmetropolitan areas, and having 3 or more comorbidities.

CONCLUSIONS: Prescription caps were associated with reduced continuity of MOUD, overall and for buprenorphine and methadone, suggesting that revising or removing these policies could improve OUD quality metrics in Medicaid.