When the machine is wrong - Characteristics of true and false predictions of Out-of-Hospital Cardiac arrests in emergency calls using a machine-learning model

Aug 4, 2026

A machine-learning model trained to identify Out-of-Hospital Cardiac Arrest (OHCA) was evaluated in real-world emergency calls to Copenhagen Emergency Medical Services between September 2018 and December 2019. The study examined call characteristics when the model failed to recognize OHCA or produced false OHCA interpretations. Emergency calls were linked to dispatch data, verified cases were obtained via the Danish Cardiac Arrest Registry, and false negative/positive cases were reviewed by trained auditors using descriptive analyses. Results reported sensitivity and specificity, plus language barriers and alternative conditions driving errors.

LanguageEnglish

CategoryHealthcare

FormatPDF

Pages8

Reading Time20 min

AccessFree to download

Abstract

Background

Methods

Results

Conclusion

Keywords

University of Southern Denmark When the machine is wrong. Characteristics of true and false predictions of Out-of-Hospital Cardiac arrests in emergency calls using a machine-learning model Nikolaj Blomberg, Stig; Jensen, Theo W. ; Porsborg Andersen, Mikkel; Folke, Fredrik; Kjær Ersbøll, Annette; Torp-Petersen, Christian; Lippert, Freddy; Collatz Christensen, Helle Published in: Resuscitation DOI: 10.1016/j.resuscitation.2023.109689 Publication date: 2023 Document version: Final published version Document license: CC BY Citation for pulished version (APA): Nikolaj Blomberg, S. , Jensen, T. W. , Porsborg Andersen, M. , Folke, F. , Kjær Ersbøll, A. , Torp-Petersen, C. , Lippert, F. , & Collatz Christensen, H. (2023) . When the machine is wrong. Characteristics of true and false predictions of Out-of-Hospital Cardiac arrests in emergency calls using a machine-learning model. Resuscitation, 183, Article 109689. [https://doi.org/10.1016/j.resuscitation.2023.109689](https://doi.org/10.1016/j.resuscitation.2023.109689) Go to publication entry in University of Southern Denmark's Research Portal Terms of use This work is brought to you by the University of Southern Denmark. Unless otherwise specified it has been shared according to the terms for self-archiving. If no other license is stated, these terms apply: • You may download this work for personal use only. • You may not further distribute the material or use it for any profit-making activity or commercial gain • You may freely distribute the URL identifying this open access version If you believe that this document breaches copyright please contact us providing details and we will investigate your claim. Please direct all enquiries to [puresupport@bib.sdu.dk](puresupport@bib.sdu.dk) Download date: 03. Aug. 2026 R E S U S C I T A T I O N 183 (20 23) 1096 89 | | | | --- | --- | |
Available online at ScienceDirect
Resuscitation

journal [homepage: www.elsevier.com/locate/resuscitation](homepage: www.elsevier.com/locate/resuscitation)

| | Clinical paper When the machine is wrong. Characteristics of true and false predictions of Out-of-Hospital Cardiac arrests in emergency calls using a machinelearning model Stig Nikolaj Blomberg a,b,*, Theo W. Jensen a,b, Mikkel Porsborg Andersen f, Fredrik Folkea,b,d, Annette Kjær Ersbølla,e, Christian Torp-Petersen f,g, Freddy Lipperta,b,h, Helle Collatz Christensen a,b,c Abstract Background: A machine-learning model trained to recognize emergency calls regarding Out-of-Hospital Cardiac Arrest (OHCA) was tested in clinical practice at Copenhagen Emergency Medical Services (EMS) from September 2018 to December 2019 . We aimed to investigate emergency call characteristics where the machine-learning model failed to recognize OHCA or misinterpreted a call as being OHCA. Methods: All emergency calls were linked to the dispatch database and verified OHCAs were identified by linkage to the Danish Cardiac Arrest Registry. Calls with either false negative or false positive predictions of OHCA were evaluated by trained auditors. Descriptive analyses were performed with absolute numbers and percentages reported. Results: The machine-learning model processed 169,236 calls to Copenhagen EMS and suspected 5,811 (3.4%) of the calls as OHCA, resulting in 84.5% sensitivity and 97.1% specificity. Among OHCAs not recognised by machine-learning model, a condition completely different from OHCA was presented by caller in 31% of the cases. In 28% of unrecognised calls, patient was reported breathing normally, and language barriers were identified in 23% of the cases. Among falsely suspected OHCA, the patient was reported unconscious in 28% of the cases, and in 13% of the false positive cases the machine-learning model interpreted calls regarding dead patients with irreversible signs of death as OHCA. Conclusion: Continuous optimization of the language model is needed to improve the predi

FAQs

What was the study aiming to investigate about the machine-learning model?

The study aimed to investigate emergency call characteristics in which the model either failed to recognize OHCA or misinterpreted a call as OHCA.

How were verified OHCA cases and false predictions assessed?

Emergency calls were linked to the dispatch database and verified OHCA cases were identified through linkage to the Danish Cardiac Arrest Registry, then false negative/positive predictions were evaluated by trained auditors.

What kinds of situations contributed to missed or falsely suspected OHCA?

For unrecognized OHCA, callers often described conditions different from OHCA, including normal breathing and identified language barriers. For falsely suspected OHCA, patients were reported unconscious in some cases, and in a subset the model interpreted irreversible signs of death as OHCA.

Aug 4, 2026

Upload by Evangeline