AI-Based Delirium Risk Prediction Compared With the Validated 4AT-Score Screening Tool in Patients With Trauma: Prospective Cohort Study
KI-basierte Delir-Risikovorhersage im Vergleich zum validierten 4AT-Score-Screeningtool bei Patient:innen mit Trauma: prospektive Kohortenstudie
JMIR Formative Research · 2026-10-21 · 10 · e104907-e104907
DOI: 10.2196/104907 ↗Auf einen Blick
Bei 2.002 Trauma-Patient:innen sagte ein KI-Tool das spätere Delir häufiger frühzeitig voraus als der 4AT-Score. Die Spezifität war beim 4AT höher.
Fragestellung
Vergleich von 4AT-Score und einem KI-gestützten Prognosetool zur Delir-Risikovorhersage in der klinischen Praxis bei Trauma-Patient:innen.
Methode
Prospektive Kohortenstudie mit 2.002 Patient:innen aus einer Trauma-Station; Risikobeurteilung bei Aufnahme, am Abend der Aufnahme und am Folgetag; KI mit random forest-Modell aus EHR-Daten.
Ergebnisse
Von 189 Patient:innen mit Delirbeginn wurden 87,3 % (165/189) durch das KI-Tool als gefährdet erkannt, gegenüber 28,0 % (5/189) mit dem 4AT. Ohne Delir wurden 73,9 % (1340/1813) vs. 96,3 % (1746/1813) korrekt als nicht gefährdet klassifiziert; AUC 0,884 (95 %-KI 0,8634–0,9039).
Schlussfolgerung der Autor:innen
Das KI-basierte Delir-Prognosetool zeigte eine bessere frühe Vorhersagefähigkeit für das Delirrisiko bei Aufnahme als der 4AT-Score.
Kernaussagen
- Frühe Delir-Vorhersage: 87,3 % mit KI vs. 28,0 % mit 4AT
- AUC 0,884 (95 %-KI 0,8634–0,9039)
- 4AT klassifizierte Nicht-Delir häufiger korrekt als das KI-Tool
Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.
At a glance
Vergleichsstudie in JMIR Formative Research, published online on 2026-10-21. Written by Klaus Franceschini and 10 co-authors. Institution: Department of Orthopaedics and Trauma, Medical University of Graz, Graz, Styria, Austria, 43 31638581719. Major MEDLINE topics: Delirium, Mass Screening, Wounds and Injuries, Artificial Intelligence. 31 references. The full text is freely available in PubMed Central.
Topics: Delir
Abstract (original language)
Background
Delirium is a complex neuropsychiatric syndrome whose severe and lasting consequences in hospitalized patients are largely considered preventable. Prevention primarily relies on the early identification of high-risk patients prior to syndrome onset. However, this remains challenging in clinical practice due to the absence of a diagnostic gold standard and the considerable administrative burden associated with established screening tools such as the 4AT score.
Objective
The aim of this study is to compare the 4AT score with an established AI-supported prognostic tool in applied clinical practice. Our goal is to demonstrate that shifting from a purely diagnostic approach toward AI-supported delirium risk prediction enables more efficient clinical workflows, reduces misclassification rates, facilitates earlier and more targeted interventions, and ultimately improves long-term patient outcomes.
Methods
In this prospective study, a total of 2002 patients from a trauma ward were screened and assessed for delirium risk using the 4AT score and the AI-based tool at admission, on the evening of admission, and on the following day, with each assessment predicting the risk of delirium occurrence at any point during the remainder of the hospital stay. The AI tool based its predictions on clinical data automatically derived from electronic health records. The AI-based tool applied a previously developed and validated random forest model (not developed as part of this study).
Results
Of the 2002 processed patients, the AI prediction tool correctly predicted 87.3% (165/189) of patients with delirium onset as at-risk, compared to an early detection rate of 28.0% (5/189) achieved by the 4AT score. Moreover, the AI-based tool correctly classified 73.9% (1340/1813) patients without delirium as not-at-risk, compared to 96.3% (1746/1813) for the 4AT score. Therefore, the AI-based tool achieved very good delirium predictions, with an area under the receiver operating characteristic curve of 0.884 (95% CI 0.8634‐0.9039).
Conclusions
The AI-based delirium prognostic tool demonstrated superior early predictive capacity for delirium risk at admission, compared with the 4AT score. The AI-based tool enables proactive risk management and targeted risk reduction.
Subjects
MeSH terms (MEDLINE)
Author keywords
Authors and institutions
- Department of Orthopaedics and Trauma, Medical University of Graz, Graz, Styria, Austria, 43 31638581719.
- Directorate for Digitalisation, IT, Organisational Development and Innovations, Styrian Hospital Holding, Graz, Austria.
- Department of Psychiatry, Psychosomatics and Psychotherapeutic Medicine, Medical University of Graz, Graz, Austria.
- Department of Neurology, Medical University of Graz, Graz, Austria.
- Hospital Pharmacy, University Hospital of Graz, Styrian Hospital Holding, Graz, Austria.
- Research Unit for Safety and Sustainability in Health Care, c/o Division of Plastic, Aesthetic and Reconstructive Surgery, Department of Surgery, Medical University of Graz, Graz, Austria.
Cite
APA
Franceschini, K., Valentini, M., Leithner, A., Kramer, D., Jauk, S., Lorenzer, L. D., Hasiba-Seebacher, K., Sumerauer, S., Godschachner, V. C., Hellauer, M., & Sendlhofer, G. (2026). AI-Based Delirium Risk Prediction Compared With the Validated 4AT-Score Screening Tool in Patients With Trauma: Prospective Cohort Study. JMIR Formative Research, 10, e104907-e104907. https://doi.org/10.2196/104907
Vancouver
Franceschini K, Valentini M, Leithner A, Kramer D, Jauk S, Lorenzer LD, et al. AI-Based Delirium Risk Prediction Compared With the Validated 4AT-Score Screening Tool in Patients With Trauma: Prospective Cohort Study. JMIR Formative Research. 2026;10:e104907-e104907. doi:10.2196/104907
BibTeX
@article{Franceschini2026,
author = {Franceschini, Klaus and Valentini, Marisa and Leithner, Andreas and Kramer, Diether and Jauk, Stefanie and Lorenzer, Lena Delia and Hasiba-Seebacher, Karin and Sumerauer, Stefan and Godschachner, Viktoria Christin and Hellauer, Magdalena and Sendlhofer, Gerald},
title = {AI-Based Delirium Risk Prediction Compared With the Validated 4AT-Score Screening Tool in Patients With Trauma: Prospective Cohort Study},
journal = {JMIR Formative Research},
year = {2026},
volume = {10},
pages = {e104907--e104907},
doi = {10.2196/104907}
}
EUVIMED