DWDeborah J. White

Deborah J. White

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article


Deborah J. White has a total of 13 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Policy-Anchored Evaluation of Large Language Model Disposition Recommendations for ED Chest Pain: Single-Institution Feasibility Study (RAPID-ED Phase 0) (Preprint)

    by Carlo Lutz, Alan Teigman, Jennifer Zapata, Kyle Shibuya, Deborah White, Benjamin Friedman, Michael Jones, Shitij Arora

    BACKGROUND Large language models (LLMs) are increasingly deployed as clinical decision support systems (CDSS), but the accuracy of their emergency department (ED) disposition recommendations is unclear. We propose measuring not just accuracy but also direction of error and adherence to institutional protocols. OBJECTIVE To assess the accuracy of policy-anchored LLM disposition recommendations for chest pain presentations to the ED, and to characterize concordance, direction of discordance, and institutional disposition-vocabulary adherence across 5 artificial intelligence (AI) configurations. METHODS In a single-institution feasibility study, 31 ED chest pain cases (21 real, 10 synthetic borderline-risk vignettes) were scored on a 5-tier ordinal disposition scale (discharge, observation, floor, telemetry, intensive care/cath lab). An emergency physician (EP) adjudicating against the institutional chest pain policy served as the reference standard (Reviewer 1) and was compared to a second EP’s review to establish ceiling inter-rater reliability. Five AI configurations were asked to provide level of care and campus disposition: an ungrounded general-purpose model (ChatGPT-Edu, GPT-5.5); 3 retrieval-augmented generation (RAG) -like configurations grounded in the 2021 national chest pain guideline and institutional disposition policy, differing in knowledge-delivery prompting; and OpenEvidence, a literature-grounded RAG LLM. The primary metric was linear-weighted Gwet AC2, with 95% CIs by bootstrap. Discordant cases were classified as over- or under-triage; agreement with actual disposition was also measured. RESULTS Physician adjudication agreed on 30/31 cases; inter-rater reliability was AC2 0.98, 95% CI 0.95-1.00. Among LLM configurations, per-case in-chat delivery achieved the highest agreement (AC2 0.53, 95% CI 0.39-0.68) and single-batch delivery the lowest (0.45), with OpenEvidence at 0.49; the ungrounded baseline was lowest overall (0.31, 95% CI 0.10-0.52). Reviewer 1 agreed with actual disposition (n=21) at AC2 0.81 (0.64-0.96), exceeding every configuration. GPT base over-triaged (22 over vs 3 under; telemetry or intensive care in 77% of cases vs 26% for the reference), whereas RAG-batch and OpenEvidence tended to under-triage (6 over vs 14 under). CONCLUSIONS No LLM configuration approached the level of physician-physician agreement. Institutional grounding improved concordance but reversed the direction of residual error, from over-triage to under-triage. This error was driven predominantly by an observation pathway not actually present at the institution. Adherence to the institutional disposition vocabulary is a distinct evaluation axis from accuracy; deployment should constrain model output to the local disposition menu.

  • Open Access

    Use of clinical phenotypes to characterize emergency department patients administered intravenous opioids for acute pain

    by Mordechai Caplan, Benjamin W. Friedman, Jason Siebert, Mai Takematsu, Victoria Adewunmi, Chiraag Gupta, Deborah J. White, Eddie Irizarry

    Objective Individual experience with opioids is highly variable. Some patients with acute pain do not experience pain relief with opioids, and many report no euphoria or dysphoric reactions. In this study, we describe the clinical phenotypes of patients who receive intravenous opioids.Methods This was an emergency department-based study in which we enrolled patients who received an intravenous opioid. We collected 0 to 10 pain scores prior to opioid administration and 15 minutes after. We also used 0 to 10 instruments to determine how high and how much euphoria the patient felt after receipt of the opioid. Using a cutoff point of ≥50% improvement in pain and the median score on the high and euphoria scales, we assigned each participant to one of the following clinical phenotypes: pain relief with feeling high or euphoria, pain relief without feeling high or euphoria, inadequate relief with feeling high or euphoria, and inadequate relief without feeling high or euphoria.Results A total of 713 patients were enrolled, 409 (57%) of whom reported not feeling high, and 465 (65%) reported no feeling of euphoria. Median percent improvement in pain was 37.5% (interquartile range, 12.5%–60.0%). One hundred seventy-eight participants (25%) were classified as experiencing pain relief with euphoria or feeling high, 190 (27%) experienced inadequate relief with euphoria or feeling high, 101 (14%) experienced pain relief without euphoria or feeling high, and 244 (34%) reported inadequate relief without euphoria or feeling high.Conclusion Among patients who receive intravenous opioids in the emergency department, the experiences of pain relief and euphoria are highly variable. For many, pain relief is independent of feeling high.