Regenstrief Institute

BackgroundOpioid use disorder (OUD) remains a critical public health crisis in the United States. Despite widespread policy and clinical interventions, early identification of individuals at risk for developing OUD remains challenging due to limitations in traditional screening approaches and a lack of individualized risk stratification methods. Machine learning (ML) methods offer an opportunity to develop timely, high-performing, and explainable predictive models that can enhance OUD prevention strategies in clinical settings.ObjectiveThis study aims to develop and validate an ML model using electronic health record (EHR) data to predict the 3-month risk of incident OUD among adults initiating opioid therapy and to stratify patients into clinically actionable risk groups.MethodsThis prognostic modeling study used 2017‐2022 OneFlorida+ EHR data to develop and validate ML algorithms predicting 3-month incident OUD risk. We included 182,083 adults (≥18 y) without cancer, overdose, or OUD or hospice history who received ≥1 outpatient, noninjectable opioid prescription. Using 183 predictors measured in sequential 3-month intervals, we developed an elastic net, least absolute shrinkage and selection operator, gradient boosting machine (GBM), and random forest models on randomly split training, testing, and validation sets. Model performance was assessed using C-statistics, predictive values, and number needed to evaluate, with patients stratified into risk deciles for clinical applicability. Model explainability was assessed using Shapley additive explanations, and fairness was evaluated using standard metrics. We externally validated the best-performing model using an independent cohort from the 2018‐2020 UPMC (formerly University of Pittsburgh Medical Center) health system.ResultsIn the validation sample (n=60,694), GBM (C-statistics=0.879, 95% CI 0.874‐0.884) and elastic net (C-statistics=0.872, 95% CI 0.867‐0.877) outperformed least absolute shrinkage and selection operator (C-statistics=0.846, 95% CI 0.840‐0.851) and random forest (C-statistics=0.798, 95% CI 0.792‐0.804), with GBM model requiring the fewest predictors (n=75) for predicting 3-month incident OUD. Using the GBM algorithm to predict the subsequent 3-month OUD risk, the top decile subgroup had a positive predictive value of 3.26%, a negative predictive value of 99.8%, and a number needed to evaluate of 31. The top decile (n=6696) captured ~68% of patients with OUD. Shapley additive explanations analysis identified age, number of outpatient visits, history of back and other pain conditions, comorbidity burden, and opioid prescribing patterns as the strongest predictors of incident OUD. Fairness assessment showed an acceptable false negative rate parity across race, age, and sex. In external validation on the UPMC cohort, the GBM model maintained good discrimination (C-statistics=0.756, 95% CI 0.750‐0.762) and effective risk stratification.ConclusionsAn ML algorithm predicting incident OUD derived from OneFlorida+ EHR data performed well in external validation with data using UPMC. The algorithm might be valuable for incident OUD risk prediction and stratification across health systems, with potential to inform early intervention.

Patients with decompensated cirrhosis have a high symptom burden and poor outcomes. Collaborative care models that provide coordinated, personalized care could improve outcomes. In this pilot randomized trial, we tested a cirrhosis-centric collaborative care model: the Cirrhosis Medical Home (CMH). A single-center pilot randomized trial enrolling 40 hospitalized adults with decompensated cirrhosis randomized 1:1 to CMH or usual care. CMH involved 6 months of postdischarge individualized care. Primary outcomes were feasibility-based (enrollment, retention, data completeness). Secondary outcomes at 3 and 6 months included quality of life (SF-36) and healthcare utilization. Of 205 patients screened, 40 were enrolled. Of the 20 patients randomized to CMH, 13 received postdischarge CMH follow-up. At 3 months, 19 died or had a transplant, 9 were lost to follow-up, and 12 completed the SF-36. At 6 months, an additional 4 died or were lost to follow-up, and 8 completed the SF-36. In intention-to-treat analysis at 3 months, CMH did not improve quality of life. In per protocol analysis at 3 months, CMH improved physical functioning (delta +15 vs -10, P = 0.015), energy/fatigue (+20 vs -5, P = 0.02), and physical component score (+5.5 vs -5.0, P = 0.009). High mortality and readmission rates were seen in both arms but without significant differences. Enrollment and retention in a randomized trial of the CMH for posthospitalization management of decompensated cirrhosis is challenging, and patient-reported outcome assessment is limited by high rates of mortality, transplant, and loss to follow-up. These data can be used to inform future design and testing of health services interventions for this population.

Delirium, an acute and fluctuating neurocognitive disorder prevalent among hospitalized and geriatric surgical patients, remains a pervasive yet underrecognized clinical challenge. Leveraging Electronic Health Records (EHRs), Machine Learning (ML) models have emerged as promising tools for early prediction and intervention. This scoping review synthesizes the existing literature, identifies current research gaps, and outlines future directions to advance delirium prediction modeling. Following the PRISMA Extension for Scoping Reviews (PRISMA-ScR) guidelines, literature from 2020 to 2025 was systematically searched across Google Scholar, EMBASE, PubMed, Scopus, and Web of Science using a comprehensive query strategy. The review highlights a significant reliance on structured preoperative and intraoperative EHR for delirium prediction, despite the existence of abundant and highly informative unstructured clinical narratives. Furthermore, a substantial heterogeneity exists in the utilized delirium identification methodologies (e.g. Nursing Delirium Screening Scale (Nu-DESC), Delirium Observation Screening Scale (DOSS), International Classification of Diseases (ICD) criteria, 4AT delirium detection, Confusion Assessment Method (CAM), Intensive Care Delirium Screening Checklist (ICDSC), Cornell Assessment of Pediatric Delirium (CAPD) Diagnostic and Statistical Manual of Mental Disorders 5th version (DSM-5), natural language processing (NLP) based analysis), alongside a focus on specific surgical subgroups. This limited data utilization and methodological variation pose challenges to ML model generalizability and robustness. The literature also showed a research emphasis on critically ill patients, potentially overlooking subtle delirium in low-severity cases. Future research should focus on early risk stratification and prioritize four key areas: (1) expanded utilization of both tabular EHR and unstructured clinical notes; (2) development of integrated multimodal fusion models adaptable to dynamic patient states; (3) investigation of the temporal dynamics of delirium development using time-series analysis; and (4) application of causal inference methods to elucidate the relationships between risk factors and delirium. Superior prediction performance can be achieved by leveraging cutting-edge architectures (e.g. transformers) and parallel computing efficiencies to move beyond traditional machine learning. To enhance real-world adoption, future work should integrate Explainable AI tools such as Shapley Additive Explanations (SHAP) within EHR-based decision support systems, improving interpretability and mitigating subgroup disparities in localized risk assessment.

Black women in early midlife experience disproportionate exposure to stress and elevated cardiovascular risk, including hypertension. Despite this, few stress management interventions are designed with and for this population. This study aims to explore the lived experiences of stress and coping among black women in early midlife with elevated blood pressure to inform the codesign of a culturally relevant, multilevel stress management intervention. We will conduct one-time, semistructured focus groups with black women aged 35-50 who have elevated blood pressure, recruited from a large safety-net health system. Data will be analysed using a constructivist grounded theory approach, with inductive theme development supported by line-by-line, focused and theoretical coding. The Social Determinants of Cardiovascular Disease framework will serve as a sensitising guide to multilevel contextual factors rather than a prescriptive coding structure. An artificial intelligence (AI)-assisted analytic component will complement human-led analysis by supporting preliminary theme exploration and enhancing transparency. Approved by the Indiana University Institutional Review Board (Protocol #21785). All participants will provide written informed consent. Findings will be shared via peer-reviewed publications, conference presentations and lay summaries for stakeholders.

Delirium is a neuropsychiatric syndrome characterized by fluctuating disturbances in attention and awareness, associated with worse clinical outcomes and higher mortality. Previous research studies have noted an association between geographic disadvantage and delirium, but it is unknown if this association extends to critically ill adults. This study aimed to explore the relationship between geographic disadvantage and ICU delirium characteristics. We performed a secondary analysis of data collected from an National Institutes of Health-funded clinical trial, the Pharmacologic Management of Delirium study. Adults 18 years old or older admitted to the ICU who experienced delirium based on the Confusion Assessment Method for the ICU (CAM-ICU) were included. None. The study population included 326 participants: 54.5% were female and 48% Black, with a mean age of 60.3 years, mean Acute Physiology and Chronic Health Evaluation II score of 20, and in-hospital mortality rate of 12.3%. The area deprivation index (ADI), a composite measure of geographic disadvantage derived from census data that yields a national percentile score ranging from 1 to 100 (with higher scores representing greater disadvantage), was obtained for each participant's address. Main outcome variables included delirium duration, which was assessed by the number of delirium- and coma-free days (DCFDs), and delirium severity, which was assessed by mean CAM-ICU-7 scores. Analysis of covariance models were used to examine differences in DCFDs and mean CAM-ICU-7 scores between ADI quintiles while controlling for demographic and clinical variables. Other clinical outcomes of interest included discharge home rates and in-hospital mortality. The sample was heavily skewed toward higher national ADI percentile scores (indicating greater disadvantage); only 11.7% of patients had an ADI score lower than 50. Our regression analyses did not reveal any associations between ADI quintile and DCFDs or mean CAM-ICU-7 scores, or between ADI quintile and discharge home rates or in-hospital mortality. However, the Black race was associated with longer delirium duration and greater delirium severity in the first week of ICU hospitalization. Our study did not find an association between geographic disadvantage and delirium duration or severity in the ICU. However, an association with race was observed, highlighting the need for further research into how socioeconomic determinants of health relate to delirium.

Background and ObjectivesThere is a need to understand Black older adults’ perceptions and attitudes about commercial activity trackers to measure and monitor outcomes in clinical trials. We sought to identify the preferred activity tracker features of Black older adults living with memory challenges or dementia and their care partners.MethodsUtilizing a mixed-methods convergent parallel design, 9 participants were recruited from Eskenazi Health in Indianapolis, Indiana. Data were collected through 2 focus groups with participants (n = 3) and care partners (n = 4), and a group interview with 1 participant and 2 care partners. The focus groups were guided by semi-structured interviews, whereas participants interacted with 4 common consumer activity tracking devices (Fitbit Inspire 3, Apple Watch SE, Polar Watch, Oura Ring Heritage). Audio recordings were analyzed using the Rapid Identification of Themes from Audio Recordings method. Participants ranked each device based on comfort, convenience, and features (eg, tracked outcomes of activity, distance/GPS, and respiratory rate). Device rankings were summarized with descriptive statistics.ResultsParticipants with memory challenges rated Apple Watch SE highest, with mean scores in comfort (4.3), convenience (3.3), and features (4.3). Care partners rated Fitbit Inspire 3 highest in comfort and Apple Watch SE for convenience and features. Qualitative findings highlighted physical attributes and comfort (large screen size), convenience (viewing progress), and features (having an emergency button and GPS).Discussion and ImplicationsFindings can guide the selection of activity trackers in future research for this population and may increase wear time and adherence in clinical trials.

BackgroundOpioids are a widely prescribed class of medication for pain management. However, they have variable efficacy and adverse effects among patients, due to the complex interplay between biological and clinical factors. Pharmacogenetic testing can be used to match patients’ genetic profiles to individualize opioid therapy, improving pain relief and reducing the risk of adverse effects. Despite its potential, the pharmacogenetic testing uptake (use of pharmacogenetic testing) remains low due to a range of barriers at the patient, health care provider, infrastructure, and financial levels. Since testing typically involves a shared decision between the provider and patient, predicting the likelihood of a patient undergoing pharmacogenetic testing and understanding the factors influencing that decision can help optimize resource use and improve outcomes in pain management.ObjectiveThis study aimed to develop machine learning (ML) models, identifying patients’ likelihood of pharmacogenetic uptake based on their demographics, clinical variables, medication use, and social determinants of health.MethodsWe used electronic health record data from a single center health care system to identify patients prescribed opioids. We extracted patients’ demographics, clinical variables, medication use, and social determinants of health, and developed and validated ML models, including a neural network, logistic regression, random forest, extreme gradient boosting (XGB), naïve Bayes, and support vector machines for pharmacogenetic testing uptake prediction based on procedure codes. We performed 5-fold cross-validation and created an ensemble probability-based classifier using the best-performing ML models for pharmacogenetic testing uptake prediction. Various performance metrics, uptake stratification analysis, and feature importance analysis were used to evaluate the performance of the models.ResultsThe ensemble model using XGB and support vector machine–radial basis function classifiers had the highest C-statistics at 79.61%, followed by XGB (78.94%), and neural network (78.05%). While XGB was the best-performing model, the ensemble model achieved a high accuracy (32,699/48,528, 67.38%), recall (537/702, 76.50%), specificity (32,162/47,826, 67.25%), and negative predictive value (32,162/32,327, 99.49%). The uptake stratification analysis using the ensemble model indicated that it can effectively distinguish across uptake probability deciles, where those in the higher strata are more likely to undergo pharmacogenetic testing in the real world (320/4853, 6.59% in the highest decile compared to 6/4853, 0.12% in the lowest). Furthermore, Shapley Additive Explanations value analysis using the XGB model indicated age, hypertension, and household income as the most influential factors for pharmacogenetic testing uptake prediction.ConclusionsThe proposed ensemble model demonstrated a high performance in pharmacogenetic testing uptake prediction among patients using opioids for pain. This model can be used as a decision support tool, assisting clinicians in identifying patients’ likelihood of pharmacogenetic testing uptake and guiding appropriate decision-making.