Arjun K. Venkatesh

Objectives: Healthcare Effectiveness Data and Information Set (HEDIS) measure Follow-up After Emergency Department Visit for Substance Use (FUA) assesses care coordination after discharge from an emergency department visit for a high-risk substance-use event. We hypothesized that the FUA denominator undercounts OUD-related ED visits. Methods: We audited a multihospital public health system’s electronic health record (EHR) to identify ED visits by patients with OUD to compare these with the OUD visits captured in the system’s automated FUA reporting. Results: From April 1, 2024, through February 28, 2025, there were 1211 emergency department (ED) unique visits related to opioid use disorder (OUD) or overdose. Only 239 (19.7%) of OUD/overdose encounters were using HEDIS methods for FUA. In multivariable logistic regression, ED visits at one of the hospitals were less likely to be included in the FUA measure (adjusted odds ratio, 0.6, 95% CI: 0.4–0.9). Conclusions: We found the FUA measure failed to identify 4 out of 5 ED visits involving high-risk opioid use disorder–related events. This suggests FUA may obscure quality performance and introduces variability into the national quality measure.

The Centers for Medicare & Medicaid Services recently introduced 2 clinical quality measures that intend, in part, to disincentivize hospital boarding: a patient safety concern that involves holding admitted patients in the emergency department while an inpatient bed is unavailable. The Age-Friendly Hospital Measure (AFHM), a structural measure, asks hospitals to attest to having protocols in place to reduce boarding for older patients. The Emergency Care Access and Timeliness (ECAT) measure, an intermediate outcome measure, will require hospitals to report on the proportion of boarded patients. Each measure was developed under a different parent quality reporting program—1 inpatient and 1 outpatient—which has the unintended consequence of producing meaningful variation between the measures. We describe opportunities to standardize and synergize the AFHM and ECAT to capture the prevalence of boarding uniformly, prevent gamesmanship, and enhance their impact on hospital performance. In particular, we highlight the importance of stratifying ECAT by the older-adult patient population to complement AFHM and since older adults are uniquely vulnerable to boarding and its associated safety risks.

Importance: Abdominal pain causes roughly 10 million US emergency department (ED) visits annually, most resulting in discharge. Post-discharge courses vary, yet existing risk models predict only whether an ED revisit occurs, not what that revisit outcome will entail. Objective: To evaluate whether Curiosity, a generative medical event foundation model, can predict post-ED-discharge trajectories for adults with abdominal pain, differentiating the timing and severity of expected outcomes. Design: Retrospective cohort study; encounters January 1–December 31, 2022; 30-day follow-up; analysis conducted in 2026. Setting: Epic Cosmos research network (multicenter, population-based, de-identified electronic health record). Participants: Adults (≥ 18 years) discharged from the ED with abdominal pain, excluding training-set patients. Random sample of 3,000 drawn from 150,030 eligible patients (65.3% female; median age 47 years [IQR 36–60]). Exposure: ED discharge after evaluation for abdominal pain. Main Outcomes and Measures: Primary: Curiosity model vs. per-task, separately estimated XGBoost models on area under the receiver operating characteristic curve (AUROC) for ED revisit ending in admission (admit-revisit), ED revisit ending in discharge (DC-revisit), and any ED revisit at 72 hours, 7 days, and 30 days. Secondary: trajectory-level accuracy across 36 trajectory classes and edit distance vs XGBoost; calibration of simulated vs observed conditional path probabilities across 45 transitions. Results: Curiosity identified patients at high risk of revisit requiring admission more accurately than XGBoost and differentiated those likely to revisit without admission. Among 3,000 patients, Curiosity’s 30-day admit-revisit AUROC was 0.83 (95% CI 0.79–0.87) vs 0.70 (95% CI 0.65–0.75) for XGBoost (DeLong P<.001), and admit-revisit AUC-PR was 0.37 (95% CI 0.29–0.46) against a 4.1% cohort base rate, vs XGBoost 0.13 (95% CI 0.09–0.19). Curiosity identified the most likely trajectory out of 36 possibilities for 45.9% of patients (XGBoost 41.0%; McNemar P<.001), with median edit distance 1.28 vs 1.40 (Wilcoxon P<.001). Median absolute calibration error across 45 transitions was 1.30 percentage points (95% CI 0.32–2.49). Conclusions and Relevance: A generative medical event foundation model produced calibrated trajectory-level predictions and discriminated admit-revisits more effectively than task-specific XGBoost baselines, separating patients that revisited and were admitted from those who revisited and were discharged.

The United States faces a predicted shortage of 36 500 physicians by 2036, with an increasing proportion of physicians leaving clinical practice or expressing an intent to do so. Evidence is limited about the extent to which stated intent to leave clinical practice translates to actual attrition from clinical practice and which factors are associated with this outcome. To characterize rates of physician attrition from clinical practice and to describe the factors associated with a differential likelihood of attrition. Nationwide, longitudinal study. All clinical settings. Physicians who provided care to Medicare patients between 2013 and 2022. Attrition from clinical practice. The sample consisted of 712 395 physicians (70.8% male, 90.8% in urban settings). Unadjusted rates of clinical practice attrition increased significantly from 3.5% in 2013 to 4.9% in 2019 (rate difference, 1.4 percentage points [95% CI, 1.3 to 1.4 percentage points]). Rates of attrition increased across the study period for both male and female physicians, in both rural and urban settings, across specialties, across geographic regions, and in all groups of physicians aged 35 years or older. In adjusted models, female physicians and those practicing in rural areas were more likely to leave clinical practice. Caring for Medicare beneficiaries with a greater average risk score, a greater average age, and a greater percentage of dual-eligible beneficiaries was also associated with attrition. Focus on services provided to Medicare fee-for-service beneficiaries. Rates of clinical practice attrition have increased since 2013. Specific physician and patient characteristics are associated with a greater risk for attrition. These findings have implications for workforce planning and the design of interventions to sustain the physician workforce. The Physicians Foundation.

Importance The clinical workforce composition in U.S. hospitals is shifting, with advanced practice practitioners (APPs) – nurse practitioners and physician assistants –assuming larger roles in inpatient care. How this mix relates to patient mortality is unclear. Objective To investigate the association between hospital-level physician proportion and 30-day risk standardized mortality rate (RSMRs) for six common inpatient conditions. Design Cross-sectional national study linking 2022 CMS Physician & Other Practitioners, Facility Affiliation, Hospital Compare Complications and Deaths, and American Hospital Association Annual Survey data. Analyses were completed October 17, 2025. Setting Nation-wide U.S. hospitals treating traditional Medicare beneficiaries in 2022. Participants Medicare beneficiaries hospitalized for acute myocardial infarction, Chronic Obstructive Pulmonary Disease (COPD), Coronary Artery Bypass Graft (CABG) Surgery, heart failure, pneumonia, or stroke. Exposure(s) (for observational studies) Hospital-level physician proportion, defined as the number of physicians divided by the sum of affiliated physicians and APPs. Main Outcome(s) and Measure(s) Hospital-level condition-specific 30-day RSMRs – case-mix adjusted outcome measures presented as proportions. Results Among 3,487 hospitals (mean physician proportion 79.7% [SD, 9.4%]), mean physician proportions across quartiles ranged from 67.8% (Q1; n=872; range, 25.0–73.5%) to 91.6% (Q4; n=871; range, 86.7–99.6%). Mean hospital-level RSMRs (%, 95% CI) were 12.63 (12.57–12.69) for acute myocardial infarction, 2.93 (2.88–2.99) for CABG surgery, 9.50 (9.44–9.56) for COPD, 12.03 (11.96–12.11) for heart failure, 18.12 (18.02–18.21) for pneumonia, and 13.74 (13.66–13.82) for stroke. Hospitals in the highest physician proportion quartile (Q4) had lower RSMRs than hospitals [p value, 95% CI] in the lowest quartile (Q1) for all conditions – acute myocardial infarction (12.54 vs. 12.86 [pConclusions and Relevance Higher relative physician staffing was modestly but consistently associated with lower hospital-level mortality across common and high-burden medical conditions. Workforce composition may be a key structural determinant of hospital quality and warrants consideration in workforce and quality improvement strategies.