EUVIMEDEUVIMED entdecken
Open Access

Detecting and Remediating Harmful Data Shifts for the Responsible Deployment of Clinical AI Models

Vallijah Subasri, Amrit Krishnan, Ali Kore, Azra Dhalla, Deval Pandya, Bo Wang, David Malkin, Fahad Razak, Amol A Verma, Anna Goldenberg, Elham Dolatabadi

JAMA Network Open · 04.06.2025 · 8 (6) · e2513685

DOI: 10.1001/jamanetworkopen.2025.13685 ↗

Kurz gefasst

Fachartikel in der Zeitschrift JAMA Network Open, online erschienen am 04.06.2025. Verfasst von Vallijah Subasri und 10 weiteren Autor:innen. Beteiligte Einrichtung: Peter Munk Cardiac Centre, University Health Network, Toronto, Ontario, Canada. Hauptthemen laut MEDLINE: COVID-19, Hospital Mortality, Artificial Intelligence. Die Arbeit wurde bisher 103-mal zitiert (Crossref) und nennt 76 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: CIHR.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Importance Clinical artificial intelligence (AI) systems are susceptible to performance degradation due to data shifts, which can lead to erroneous predictions and potential patient harm. Proactively detecting and mitigating these shifts is crucial for maintaining AI effectiveness and safety in clinical practice.

Objectives

To develop and evaluate a proactive, label-agnostic monitoring pipeline to detect and mitigate harmful data shifts in clinical AI systems and to assess the use of transfer learning and continual learning strategies in maintaining model performance. Design, Setting, and Participants This prognostic study was conducted using electronic health record data for admissions to general internal medicine wards of 7 large hospitals (5 academic and 2 community) in Toronto, Canada, between January 1, 2010, to August 31, 2020. Inpatients (aged ≥18 years) with a hospital stay of at least 24 hours were included. Data analysis was performed from January to August 2022. Exposures Data shifts due to changes in hospital type, critical laboratory assays, patient demographics, admission type, and the COVID-19 pandemic. Main Outcomes and Measures The primary outcome was predictive performance for all-cause in-hospital mortality within the next 2 weeks, evaluated using the area under the receiver operating characteristic curve (AUROC) and the area under the precision-recall curve (AUPRC). Data shifts were detected using a label-agnostic monitoring pipeline employing a black box shift estimator with maximum mean discrepancy testing.

Results

Data were available for 143 049 adult inpatients (mean [SD] age, 67.8 [19.6] years; 50.7% female). Significant data shifts were detected as a result of changes in younger age groups and admissions from nursing homes and acute care centers, transferring from community to academic hospitals, and changes in brain natriuretic peptide and D-dimer. Transfer learning improved model performance of community hospitals in a hospital type–dependent manner (Delta AUROC [SD], 0.05 [0.03]; Delta AUPRC [SD], 0.06 [0.04]). During the COVID-19 pandemic, drift-triggered continual learning improved overall model performance (Delta AUROC [SD], 0.44 [0.02]; P = .007, Mann-Whitney U test). Conclusions and Relevance In this prognostic study, a proactive, label-agnostic monitoring pipeline detected harmful data shifts for a clinical AI system predicting in-hospital mortality. Transfer learning and drift-triggered continual learning strategies mitigated performance degradation, maintaining model performance across health care settings. These findings suggest that the approach used here may ensure the robust and equitable deployment of clinical AI models. Future research should explore the generalizability of this framework across diverse clinical domains, data modalities, and longer deployment periods to further validate its effectiveness.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

COVID-19Hospital MortalityArtificial IntelligenceHumansFemaleMaleMiddle AgedAgedSARS-CoV-2Electronic Health RecordsAdultOntarioAged, 80 and over

Autor:innen und Institutionen

  • Peter Munk Cardiac Centre, University Health Network, Toronto, Ontario, Canada.
  • Vector Institute for Artificial Intelligence, Toronto, Ontario, Canada.
  • Genetics and Genome Biology Program, The Hospital for Sick Children, Toronto, Ontario, Canada.
  • Department of Computer Science, University of Toronto, Toronto, Ontario, Canada.
  • Division of Hematology/Oncology, The Hospital for Sick Children, Toronto, Ontario, Canada.
  • Department of Pediatrics, University of Toronto, Toronto, Ontario, Canada.

Förderung

  • CIHR

Zitieren

APA

Subasri, V., Krishnan, A., Kore, A., Dhalla, A., Pandya, D., Wang, B., Malkin, D., Razak, F., Verma, A. A., Goldenberg, A., & Dolatabadi, E. (2025). Detecting and Remediating Harmful Data Shifts for the Responsible Deployment of Clinical AI Models. JAMA Network Open, 8(6), e2513685. https://doi.org/10.1001/jamanetworkopen.2025.13685

Vancouver

Subasri V, Krishnan A, Kore A, Dhalla A, Pandya D, Wang B, et al. Detecting and Remediating Harmful Data Shifts for the Responsible Deployment of Clinical AI Models. JAMA Network Open. 2025;8(6):e2513685. doi:10.1001/jamanetworkopen.2025.13685

BibTeX
@article{Subasri2025,
  author = {Subasri, Vallijah and Krishnan, Amrit and Kore, Ali and Dhalla, Azra and Pandya, Deval and Wang, Bo and Malkin, David and Razak, Fahad and Verma, Amol A and Goldenberg, Anna and Dolatabadi, Elham},
  title = {Detecting and Remediating Harmful Data Shifts for the Responsible Deployment of Clinical AI Models},
  journal = {JAMA Network Open},
  year = {2025},
  volume = {8},
  number = {6},
  pages = {e2513685},
  doi = {10.1001/jamanetworkopen.2025.13685}
}
Kostenlos starten →