Liem M. Nguyen

All Publications

• Using electronic health record audit logs to identify pediatric intensive care unit teams. JAMIA open Nguyen, L. M., Sebok-Syer, S. S., Jacobson, D., Mariano, K., Grunyk, J., Burk, S., Nourkhalaj, Y., Kirk, K., Profit, J., Rose, C., Holmgren, A. J., Kannampallil, T., Tawfik, D. 2026; 9 (5): ooag126

  Abstract

  Teamwork influences patient outcomes, but electronic health records (EHRs) do not reliably capture patient care team members, limiting its study. EHR audit logs provide insights into healthcare workers' workflows and may help identify patients' care team members.To develop, validate, and compare algorithms utilizing EHR audit logs to identify patient-centric primary teams (bedside nurse, frontline clinician, and attending physician) in pediatric intensive care units (ICUs).We observed rounds for 1931 patient days (development = 678; validation = 1253) across pediatric (PICU), neonatal (NICU) and cardiovascular (CVICU) ICUs at a quaternary children's hospital and documented each patient's daytime care team. We developed 2 algorithms that leveraged EHR audit logs to identify team members: (1) clinically informed heuristics, and (2) a Longitudinal Contribution Score (LCS). Accuracy was computed for each role and algorithm. 95% confidence intervals (CIs) and algorithm comparisons (bootstrap P values) were computed via patient-day bootstrap resampling.In the development cohort (PICU), the LCS demonstrated greater accuracy than the heuristics for all roles: nurse (LCS 92.9% [95% CI, 90.9-94.8]; heuristic 67.6% [64.0-71.1]), frontline (83.3% [80.5-86.1]; 77.9% [74.6-81.0]), and attending (66.5% [62.7-70.2]; 61.9% [58.3-65.6]); all P < .005. In the validation cohort (PICU, NICU, and CVICU), only the nurse advantage was statistically replicated (LCS 91.2% [89.5-92.8], heuristic 74.2% [71.7-76.4]; P < .001); with frontline (83.0% [80.9-85.1], 81.2% [79.1-83.4]; P = .110) and attending (69.0 [66.5-71.6] vs 70.0% [67.4-72.4]; P = .326) roles showing comparable accuracy.EHR audit log data can accurately identify patients' bedside nurses in pediatric ICUs. For frontline clinicians and attending physicians whose clinical coproduction and interdependence blurs individual EHR signals, identification is less accurate and supplementary data sources may be needed. Both algorithms provide a scalable foundation for team dynamics research, although the LCS offering greater generalizability than predefined heuristics.

  View details for DOI 10.1093/jamiaopen/ooag126

  View details for PubMedID 42725307

  View details for PubMedCentralID PMC13561518

• Identifying Electronic Health Record Tasks and Activity Using Computer Vision. Applied clinical informatics Nguyen, L. M., Sinha, A., Dziorny, A., Tawfik, D. 2025

  Abstract

  Time spent in the electronic health record (EHR) is an important measure of clinical activity. Vendor-derived EHR use metrics may not correspond to actual EHR experience. Raw EHR audit logs enable customized EHR use metrics, but translating discrete timestamps to time intervals is challenging. There are insufficient data available to quantify inactivity between audit log timestamps.We propose a computer vision-based model that can 1) classify EHR tasks being performed, and identify when task changes occur, and 2) quantify active-use time using session screen recordings of EHR use. We generated 111 minutes of simulated workflow in an Epic sandbox environment for development and training and collected 86 minutes of real-world clinician session recordings for validation. The model used YOLOv8, Tesseract OCR, and a predefined dictionary to perform task classification and task change detection. We developed a frame comparison algorithm to delineate activity from inactivity and thus quantify active time. We compared the model's output of task classification, task change identification, and active time quantification against clinician annotations. We then performed a post-hoc sensitivity analysis to identify the model's accuracy when using optimal parameters.Our model classified time spent in various high-level tasks with 94% accuracy. It detected task changes with 90.6% sensitivity. Active-use quantification varied by task, with lower MAPE for tasks with clear visual changes (e.g., Results Review) and higher MAPE for tasks with subtle interactions (e.g., Note Entry). A post-hoc sensitivity analysis revealed improvement in active-use quantification with a lower threshold of inactivity than initially used.A computer vision approach to identifying tasks performed and measuring time spent in the EHR is feasible. Future work should refine task-specific thresholds and validate across diverse settings. This approach enables defining optimal context-sensitive thresholds for quantifying clinically relevant active EHR time using raw audit log data.

  View details for DOI 10.1055/a-2698-0841

  View details for PubMedID 40930500

• Predicting Primary Care Physician Burnout From Electronic Health Record Use Measures. Mayo Clinic proceedings Tawfik, D., Bayati, M., Liu, J., Nguyen, L., Sinha, A., Kannampallil, T., Shanafelt, T., Profit, J. 2024

  Abstract

  To evaluate the ability of routinely collected electronic health record (EHR) use measures to predict clinical work units at increased risk of burnout and potentially most in need of targeted interventions.In this observational study of primary care physicians, we compiled clinical workload and EHR efficiency measures, then linked these measures to 2 years of well-being surveys (using the Stanford Professional Fulfillment Index) conducted from April 1, 2019, through October 16, 2020. Physicians were grouped into training and confirmation data sets to develop predictive models for burnout. We used gradient boosting classifier and other prediction modeling algorithms to quantify the predictive performance by the area under the receiver operating characteristics curve (AUC).Of 278 invited physicians from across 60 clinics, 233 (84%) completed 396 surveys. Physicians were 67% women with a median age category of 45 to 49 years. Aggregate burnout score was in the high range (≥3.325/10) on 111 of 396 (28%) surveys. Gradient boosting classifier of EHR use measures to predict burnout achieved an AUC of 0.59 (95% CI, 0.48 to 0.77) and an area under the precision-recall curve of 0.29 (95% CI, 0.20 to 0.66). Other models' confirmation set AUCs ranged from 0.56 (random forest) to 0.66 (penalized linear regression followed by dichotomization). Among the most predictive features were physician age, team member contributions to notes, and orders placed with user-defined preferences. Clinic-level aggregate measures identified the top quartile of clinics with 56% sensitivity and 85% specificity.In a sample of primary care physicians, routinely collected EHR use measures demonstrated limited ability to predict individual burnout and moderate ability to identify high-risk clinics.

  View details for DOI 10.1016/j.mayocp.2024.01.005

  View details for PubMedID 38573301