Derivation and external validation of machine-learning models for risk stratification in chest pain with normal troponin

被引:2
作者
Fernandez-Cisnal, Agustin [1 ]
Lopez-Ayala, Pedro [2 ,3 ]
Valero, Ernesto [1 ]
Koechlin, Luca [2 ,3 ]
Catarrala, Arturo [4 ]
Boeddinghaus, Jasper [2 ,3 ]
Noceda, Jose [5 ]
Nestelberger, Thomas [2 ,3 ]
Miro, Oscar [6 ]
Julio, Nunez [1 ]
Mueller, Christian [2 ,3 ]
Sanchis, Juan [1 ]
机构
[1] Univ Valencia, Hosp Clin Univ Valencia, Ctr Invest Biomed Red Enfermedades Cardiovaculare, Inst Invest Sanitaria INCLIVA,Cardiol Dept, Valencia, Spain
[2] Univ Basel, Univ Hosp Basel, Univ Heart Ctr Basel, Cardiovasc Res Inst Basel CRIB, Basel, Switzerland
[3] Univ Basel, Univ Hosp Basel, Univ Heart Ctr Basel, Dept Cardiol, Basel, Switzerland
[4] Hosp Clin Univ Valencia, Inst Invest Sanitaria INCLIVA, Clin Biochem Dept, Valencia 46010, Spain
[5] Hosp Clin Univ Valencia, Inst Invest Sanitaria INCLIVA, Emergency Dept, Valencia 46010, Spain
[6] Hosp Clin Barcelona, Emergency Dept, Barcelona, Catalonia, Spain
关键词
Troponin; Prediction; Machine learning; Myocardial infarction; ACUTE MYOCARDIAL-INFARCTION; EMERGENCY-DEPARTMENT; SCORE; PREDICTION; ELEVATION; DIAGNOSIS; RULE; MACS; ESC;
D O I
10.1093/ehjacc/zuad089
中图分类号
R5 [内科学];
学科分类号
1002 ; 100201 ;
摘要
Aims Risk stratification of patients with chest pain and a high-sensitivity cardiac troponin T (hs-cTnT) concentration <upper reference limit (URL) is challenging. The aim of this study was to develop and externally validate clinical models for risk prediction of 90-day death or myocardial infarction in patients presenting to the emergency department with chest pain and an initial hs-cTnT concentration <URL. Methods and results Four machine-learning-based models and one logistic regression (LR) model were trained on 4075 patients (single-centre Spanish cohort) and externally validated on 3609 patients (international prospective Advantageous Predictors of Acute Coronary syndromes Evaluation cohort). Models were compared with GRACE and HEART scores and a single undetectable hs-cTnT-based strategy (u-cTn; hs-cTnT < 5 ng/L and time from symptoms onset >180 min). Probability thresholds for safe discharge were derived in the derivation cohort. The endpoint occurred in 105 (2.6%) patients in the training set and 98 (2.7%) in the external validation set. Gradient boosting full (GBf) showed the best discrimination (area under the curve = 0.808). Calibration was good for the reduced neural network and LR models. Gradient boosting full identified the highest proportion of patients for safe discharge (36.7 vs. 23.4 vs. 27.2%; GBf vs. LR vs. u-cTn, respectively) with similar safety (missed endpoint per 1000 patients: 2.2 vs. 3.5 vs. 3.1, respectively). All derived models were superior to the HEART and GRACE scores (P < 0.001). Conclusion Machine-learning and LR prediction models were superior to the HEART, GRACE, and u-cTn for risk stratification of patients with chest pain and a baseline hs-cTnT <URL. Gradient boosting full models best balanced discrimination, calibration, and efficacy, reducing the need for serial hs-cTnT determination by more than one-third. Clinical trial registration ClinicalTrials.gov number, NCT00470587, https://clinicaltrials.gov/ct2/show/NCT00470587.
引用
收藏
页码:743 / 752
页数:10
相关论文
共 35 条
  • [1] The TIMI risk score for unstable angina/non-ST elevation MI - A method for prognostication and therapeutic decision making
    Antman, EM
    Cohen, M
    Bernink, PJLM
    McCabe, CH
    Horacek, T
    Papuchis, G
    Mautner, B
    Corbalan, R
    Radley, D
    Braunwald, E
    [J]. JAMA-JOURNAL OF THE AMERICAN MEDICAL ASSOCIATION, 2000, 284 (07): : 835 - 842
  • [2] Prognostic implications of detectable cardiac troponin I below the 99th percentile in patients admitted to an emergency department without acute coronary syndrome
    Bardaji, Alfredo
    Bonet, Gil
    Carrasquer, Anna
    Gonzalez-del Hoyo, Maribel
    Dominguez, Fernando
    Sanchez, Rafael
    Boque, Carme
    Cediel, German
    [J]. CLINICAL CHEMISTRY AND LABORATORY MEDICINE, 2018, 56 (11) : 1954 - 1961
  • [3] Biondi MJ, 2021, CIRCULATION, V144
  • [4] Troponin-only Manchester Acute Coronary Syndromes (T-MACS) decision aid: single biomarker re-derivation and external validation in three cohorts
    Body, Richard
    Carlton, Edward
    Sperrin, Matthew
    Lewis, Philip S.
    Burrows, Gillian
    Carley, Simon
    McDowell, Garry
    Buchan, Iain
    Greaves, Kim
    Mackway-Jones, Kevin
    [J]. EMERGENCY MEDICINE JOURNAL, 2017, 34 (06) : 349 - +
  • [5] Early Diagnosis of Myocardial Infarction With Point-of-Care High-Sensitivity Cardiac Troponin I
    Boeddinghaus, Jasper
    Nestelberger, Thomas
    Koechlin, Luca
    Wussler, Desiree
    Lopez-Ayala, Pedro
    Walter, Joan Elias
    Troester, Valentina
    Ratmann, Paul David
    Seidel, Funda
    Zimmermann, Tobias
    Badertscher, Patrick
    Wildi, Karin
    Gimenez, Maria Rubini
    Potlukova, Eliska
    Strebel, Ivo
    Freese, Michael
    Miro, Oscar
    Javier Martin-Sanchez, F.
    Kawecki, Damian
    Keller, Dagmar I.
    Gualandro, Danielle M.
    Christ, Michael
    Twerenbold, Raphael
    Mueller, Christian
    [J]. JOURNAL OF THE AMERICAN COLLEGE OF CARDIOLOGY, 2020, 75 (10) : 1111 - 1124
  • [6] Collet JP, 2021, REV ESP CARDIOL, V74, DOI [10.1016/j.rec.2021.05.002, 10.1093/eurheartj/ehaa575]
  • [7] Quantifying the impact of different approaches for handling continuous predictors on the performance of a prognostic model
    Collins, Gary S.
    Ogundimu, Emmanuel O.
    Cook, Jonathan A.
    Le Manach, Yannick
    Altman, Douglas G.
    [J]. STATISTICS IN MEDICINE, 2016, 35 (23) : 4124 - 4135
  • [8] Collins GS, 2015, ANN INTERN MED, V162, P55, DOI [10.1136/bmj.g7594, 10.1016/j.jclinepi.2014.11.010, 10.7326/M14-0697, 10.1038/bjc.2014.639, 10.7326/M14-0698, 10.1016/j.eururo.2014.11.025, 10.1186/s12916-014-0241-z, 10.1002/bjs.9736]
  • [9] Clinical History and Detectable Troponin Concentrations below the 99th Percentile for Risk Stratification of Patients with Chest Pain and First Normal Troponin
    Fernandez-Cisnal, Agustin
    Valero, Ernesto
    Garcia-Blas, Sergio
    Pernias, Vicente
    Pozo, Adela
    Carratala, Arturo
    Gonzalez, Jessika
    Noceda, Jose
    Minana, Gema
    Nunez, Julio
    Sanchis, Juan
    [J]. JOURNAL OF CLINICAL MEDICINE, 2021, 10 (08)
  • [10] Prediction of risk of death and myocardial infarction in the six months after presentation with acute coronary syndrome: prospective multinational observational study (GRACE)
    Fox, Keith A. A.
    Dabbous, Omar H.
    Goldberg, Robert J.
    Pieper, Karen S.
    Eagle, Kim A.
    Van de Werf, Frans
    Avezum, Alvaro
    Goodman, Shaun G.
    Flather, Marcus D.
    Anderson, Frederick A., Jr.
    Granger, Christopher B.
    [J]. BMJ-BRITISH MEDICAL JOURNAL, 2006, 333 (7578): : 1091 - 1094