EUVIMEDEUVIMED entdecken
Leitlinie

Machine Learning for Myocardial Infarction Compared With Guideline-Recommended Diagnostic Pathways

Machine Learning für Myokardinfarkt im Vergleich zu leitlinienempfohlenen diagnostischen Pfaden

Jasper Boeddinghaus iD, Dimitrios Doudesis iD, Pedro Lopez-Ayala iD, Kuan Ken Lee iD, Luca Koechlin iD, Karin Wildi iD, Thomas Nestelberger iD, Raphael Borer, Òscar Miró iD, F Javier Martin-Sanchez iD, Ivo Strebel iD, Maria Rubini Giménez iD und 11 weitere

Circulation · 02.04.2024 · 149 (14) · 1090-1101

DOI: 10.1161/circulationaha.123.066917 ↗

Auf einen Blick

Bei 4.105 Patient:innen mit möglichem Myokardinfarkt identifizierte CoDE-ACS je nach Zeitpunkt mehr Personen als niedrig wahrscheinlich als die Leitlinienpfade. Die Vorhersageleistung war dabei vergleichbar.

Fragestellung

Wie schneidet CoDE-ACS zu verschiedenen Zeitpunkten der seriellen Troponinmessung ab und wie vergleicht sich das Modell mit leitlinienempfohlenen diagnostischen Pfaden?

Methode

Patient:innen mit möglichem MI ohne ST-Hebung an 12 Zentren in 5 Ländern; hs-cTnI-Messung bei 0, 1 und 2 Stunden; Vergleich von CoDE-ACS mit ESC 0/1-, ESC 0/2- und High-STEACS-Pfaden.

Ergebnisse

4.105 Patient:innen, 575 mit Typ-1-MI. Bei Aufnahme: 56 % niedrig wahrscheinlich; NPV 99,7 % (95 %-KI 99,5–99,9), Sensitivität 99,0 % (98,6–99,2). Nach 1/2 Stunden: NPV jeweils 99,7 %.

Schlussfolgerung der Autor:innen

CoDE-ACS war unabhängig vom Zeitpunkt der seriellen Troponinmessung konsistent und identifizierte mehr Patient:innen als niedrig wahrscheinlich bei vergleichbarer Leistung zu den leitlinienempfohlenen Pfaden. Prospektive Evaluation ist erforderlich.

Kernaussagen

  • 4.105 Patient:innen, 575 mit Typ-1-MI
  • NPV von CoDE-ACS bei Aufnahme 99,7 %
  • Mehr Patient:innen als niedrig wahrscheinlich als mit ESC- und High-STEACS-Pfaden

Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.

Kurz gefasst

Leitlinie in der Zeitschrift Circulation, erschienen am 02.04.2024. Verfasst von Jasper Boeddinghaus und 22 weiteren Autor:innen. Beteiligte Einrichtung: Cardiovascular Research Institute Basel (CRIB) and Department of Cardiology (J.B., P.L.-A., L.K., K.W., T.N., R.B., I.S., M.R.G., C.M.), University H…. Hauptthemen laut MEDLINE: Acute Coronary Syndrome, Myocardial Infarction. Die Arbeit wurde bisher 45-mal zitiert (Crossref) und nennt 49 Literaturangaben. Gefördert von: British Heart Foundation, Medical Research Council.

Auf einen BlickAbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

BACKGROUND: Collaboration for the Diagnosis and Evaluation of Acute Coronary Syndrome (CoDE-ACS) is a validated clinical decision support tool that uses machine learning with or without serial cardiac troponin measurements at a flexible time point to calculate the probability of myocardial infarction (MI). How CoDE-ACS performs at different time points for serial measurement and compares with guideline-recommended diagnostic pathways that rely on fixed thresholds and time points is uncertain. METHODS: Patients with possible MI without ST-segment–elevation were enrolled at 12 sites in 5 countries and underwent serial high-sensitivity cardiac troponin I concentration measurement at 0, 1, and 2 hours. Diagnostic performance of the CoDE-ACS model at each time point was determined for index type 1 MI and the effectiveness of previously validated low- and high-probability scores compared with guideline-recommended European Society of Cardiology (ESC) 0/1-hour, ESC 0/2-hour, and High-STEACS (High-Sensitivity Troponin in the Evaluation of Patients With Suspected Acute Coronary Syndrome) pathways. RESULTS: In total, 4105 patients (mean age, 61 years [interquartile range, 50–74]; 32% women) were included, among whom 575 (14%) had type 1 MI. At presentation, CoDE-ACS identified 56% of patients as low probability, with a negative predictive value and sensitivity of 99.7% (95% CI, 99.5%–99.9%) and 99.0% (98.6%–99.2%), ruling out more patients than the ESC 0-hour and High-STEACS (25% and 35%) pathways. Incorporating a second cardiac troponin measurement, CoDE-ACS identified 65% or 68% of patients as low probability at 1 or 2 hours, for an identical negative predictive value of 99.7% (99.5%–99.9%); 19% or 18% as high probability, with a positive predictive value of 64.9% (63.5%–66.4%) and 68.8% (67.3%–70.1%); and 16% or 14% as intermediate probability. In comparison, after serial measurements, the ESC 0/1-hour, ESC 0/2-hour, and High-STEACS pathways identified 49%, 53%, and 71% of patients as low risk, with a negative predictive value of 100% (99.9%–100%), 100% (99.9%–100%), and 99.7% (99.5%–99.8%); and 20%, 19%, or 29% as high risk, with a positive predictive value of 61.5% (60.0%–63.0%), 65.8% (64.3%–67.2%), and 48.3% (46.8%–49.8%), resulting in 31%, 28%, or 0%, who require further observation in the emergency department, respectively. CONCLUSIONS: CoDE-ACS performs consistently irrespective of the timing of serial cardiac troponin measurement, identifying more patients as low probability with comparable performance to guideline-recommended pathways for MI. Whether care guided by probabilities can improve the early diagnosis of MI requires prospective evaluation. REGISTRATION: URL: https://www.clinicaltrials.gov ; Unique identifier: NCT00470587.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Acute Coronary SyndromeMyocardial InfarctionHumansFemaleMiddle AgedMaleBiomarkersTroponinMachine LearningTroponin T

Stichwörter der Autor:innen

machine learningmyocardial infarctiontroponin

Substanzen

BiomarkersTroponinTroponin T

Autor:innen und Institutionen

  • Cardiovascular Research Institute Basel (CRIB) and Department of Cardiology (J.B., P.L.-A., L.K., K.W., T.N., R.B., I.S., M.R.G., C.M.), University Hospital Basel, University of Basel, Switzerland.
  • BHF/University Centre for Cardiovascular Science (J.B., D.D., K.K.L., A.B., Z.L., A.V.F., C.T., A.A., N.L.M.), University of Edinburgh, UK.
  • Usher Institute (D.D., K.K.L., A.G., N.L.M.), University of Edinburgh, UK.
  • Departments of Cardiac Surgery (L.K.), University Hospital Basel, University of Basel, Switzerland.
  • Intensive Care (K.W.), University Hospital Basel, University of Basel, Switzerland.
  • Emergency Department, Hospital Clinic, Barcelona, Catalonia, Spain (Ò.M.).

ORCID-Profile: Jasper Boeddinghaus, Dimitrios Doudesis, Pedro Lopez-Ayala, Kuan Ken Lee, Luca Koechlin, Karin Wildi, Thomas Nestelberger, Òscar Miró, F Javier Martin-Sanchez, Ivo Strebel, Maria Rubini Giménez, Dagmar I Keller, Michael Christ, Ziwen Li, Amy V Ferry, Chris Tuck, Atul Anand, Alasdair Gray, Christian Mueller

Förderung

  • British Heart Foundation
  • Medical Research Council

Zitieren

APA

Boeddinghaus, J., Doudesis, D., Lopez-Ayala, P., Lee, K. K., Koechlin, L., Wildi, K., Nestelberger, T., Borer, R., Miró, Ò., Martin-Sanchez, F. J., Strebel, I., Rubini Giménez, M., Keller, D. I., Christ, M., Bularga, A., Li, Z., Ferry, A. V., Tuck, C., Anand, A., …, & CoDE-ACS and APACE Investigators (2024). Machine Learning for Myocardial Infarction Compared With Guideline-Recommended Diagnostic Pathways. Circulation, 149(14), 1090-1101. https://doi.org/10.1161/circulationaha.123.066917

Vancouver

Boeddinghaus J, Doudesis D, Lopez-Ayala P, Lee KK, Koechlin L, Wildi K, et al. Machine Learning for Myocardial Infarction Compared With Guideline-Recommended Diagnostic Pathways. Circulation. 2024;149(14):1090-1101. doi:10.1161/circulationaha.123.066917

BibTeX
@article{Boeddinghaus2024,
  author = {Boeddinghaus, Jasper and Doudesis, Dimitrios and Lopez-Ayala, Pedro and Lee, Kuan Ken and Koechlin, Luca and Wildi, Karin and Nestelberger, Thomas and Borer, Raphael and Miró, Òscar and Martin-Sanchez, F Javier and Strebel, Ivo and Rubini Giménez, Maria and Keller, Dagmar I and Christ, Michael and Bularga, Anda and Li, Ziwen and Ferry, Amy V and Tuck, Chris and Anand, Atul and Gray, Alasdair and Mills, Nicholas L and Mueller, Christian and CoDE-ACS and APACE Investigators, CoDE-ACS and APACE Investigators},
  title = {Machine Learning for Myocardial Infarction Compared With Guideline-Recommended Diagnostic Pathways},
  journal = {Circulation},
  year = {2024},
  volume = {149},
  number = {14},
  pages = {1090--1101},
  doi = {10.1161/circulationaha.123.066917}
}
Kostenlos starten →