Jussi A Hernesniemi1,2, Shadi Mahdiani1,3, Juho A Tynkkynen1,4, Leo-Pekka Lyytikäinen1,2,5, Pashupati P Mishra5, Terho Lehtimäki1,5, Markku Eskola2, Kjell Nikus1,2, Kari Antila3, Niku Oksala1,5,6. 1. a Faculty of Medicine and Health Technology , Tampere University , Tampere , Finland. 2. b Department of Cardiology , Tays Heart Hospital, Tampere University Hospital , Tampere , Finland. 3. c VTT Technical Research Center of Finland , Tampere , Finland. 4. d Department of Radiology , Kanta-Häme Central Hospital , Hämeenlinna , Finland. 5. e Department of Clinical Chemistry, Fimlab Laboratories and Finnish Cardiovascular Research Center-Tampere, Faculty of Medicine and Health Technology , Tampere University , Tampere , Finland. 6. f Vascular and Interventional Radiology Centre , Tampere University Hospital , Tampere , Finland.
Abstract
Objective: Investigation of the clinical potential of extensive phenotype data and machine learning (ML) in the prediction of mortality in acute coronary syndrome (ACS). Methods: The value of ML and extensive clinical data was analyzed in a retrospective registry study of 9066 consecutive ACS patients (January 2007 to October 2017). Main outcome was six-month mortality. Prediction models were developed using two ML methods, logistic regression and extreme gradient boosting (xgboost). The models were fitted in training set of patients treated in 2007-2014 and 2017 (81%, n = 7344) and validated in a separate validation set of patients treated in 2015-2016 with full GRACE score data available for comparison of model accuracy (19%, n = 1722). Results: Overall, six-month mortality was 7.3% (n = 660). Several variables were found to be significantly associated with six-month mortality by both ML methods. The xgboost scored the best performance: AUC 0.890 (0.864-0.916). The AUC values for logistic regression and GRACE score were 0.867(0.837-0.897) and 0.822 (0.785-0.859), respectively. The AUC value of xgboost was better when compared to logistic regression (p = .012) and GRACE score (p < .00001). Conclusions: The use of extensive phenotype data and novel machine learning improves prediction of mortality in ACS over traditional GRACE score. KEY MESSAGES The collection of extensive cardiovascular phenotype data from electronic health records as well as from data recorded by physicians can be used highly effectively in prediction of mortality after acute coronary syndrome. Supervised machine learning methods such as logistic regression and extreme gradient boosting using extensive phenotype data significantly outperform conventional risk assessment by the current golden standard GRACE score. Integration of electronic health records and the use of supervised machine learning methods can be easily applied in a single centre level to model the risk of mortality.
Objective: Investigation of the clinical potential of extensive phenotype data and machine learning (ML) in the prediction of mortality in acute coronary syndrome (ACS). Methods: The value of ML and extensive clinical data was analyzed in a retrospective registry study of 9066 consecutive ACS patients (January 2007 to October 2017). Main outcome was six-month mortality. Prediction models were developed using two ML methods, logistic regression and extreme gradient boosting (xgboost). The models were fitted in training set of patients treated in 2007-2014 and 2017 (81%, n = 7344) and validated in a separate validation set of patients treated in 2015-2016 with full GRACE score data available for comparison of model accuracy (19%, n = 1722). Results: Overall, six-month mortality was 7.3% (n = 660). Several variables were found to be significantly associated with six-month mortality by both ML methods. The xgboost scored the best performance: AUC 0.890 (0.864-0.916). The AUC values for logistic regression and GRACE score were 0.867(0.837-0.897) and 0.822 (0.785-0.859), respectively. The AUC value of xgboost was better when compared to logistic regression (p = .012) and GRACE score (p < .00001). Conclusions: The use of extensive phenotype data and novel machine learning improves prediction of mortality in ACS over traditional GRACE score. KEY MESSAGES The collection of extensive cardiovascular phenotype data from electronic health records as well as from data recorded by physicians can be used highly effectively in prediction of mortality after acute coronary syndrome. Supervised machine learning methods such as logistic regression and extreme gradient boosting using extensive phenotype data significantly outperform conventional risk assessment by the current golden standard GRACE score. Integration of electronic health records and the use of supervised machine learning methods can be easily applied in a single centre level to model the risk of mortality.
Authors: David C Goff; Donald M Lloyd-Jones; Glen Bennett; Sean Coady; Ralph B D'Agostino; Raymond Gibbons; Philip Greenland; Daniel T Lackland; Daniel Levy; Christopher J O'Donnell; Jennifer G Robinson; J Sanford Schwartz; Susan T Shero; Sidney C Smith; Paul Sorlie; Neil J Stone; Peter W F Wilson; Harmon S Jordan; Lev Nevo; Janusz Wnek; Jeffrey L Anderson; Jonathan L Halperin; Nancy M Albert; Biykem Bozkurt; Ralph G Brindis; Lesley H Curtis; David DeMets; Judith S Hochman; Richard J Kovacs; E Magnus Ohman; Susan J Pressler; Frank W Sellke; Win-Kuang Shen; Sidney C Smith; Gordon F Tomaselli Journal: Circulation Date: 2013-11-12 Impact factor: 29.690
Authors: Christian W Hamm; Jean-Pierre Bassand; Stefan Agewall; Jeroen Bax; Eric Boersma; Hector Bueno; Pio Caso; Dariusz Dudek; Stephan Gielen; Kurt Huber; Magnus Ohman; Mark C Petrie; Frank Sonntag; Miguel Sousa Uva; Robert F Storey; William Wijns; Doron Zahger Journal: Eur Heart J Date: 2011-08-26 Impact factor: 29.983
Authors: Keith A A Fox; Omar H Dabbous; Robert J Goldberg; Karen S Pieper; Kim A Eagle; Frans Van de Werf; Alvaro Avezum; Shaun G Goodman; Marcus D Flather; Frederick A Anderson; Christopher B Granger Journal: BMJ Date: 2006-10-10
Authors: Christopher B Granger; Robert J Goldberg; Omar Dabbous; Karen S Pieper; Kim A Eagle; Christopher P Cannon; Frans Van De Werf; Alvaro Avezum; Shaun G Goodman; Marcus D Flather; Keith A A Fox Journal: Arch Intern Med Date: 2003-10-27
Authors: Harry Hemingway; Folkert W Asselbergs; John Danesh; Richard Dobson; Nikolaos Maniadakis; Aldo Maggioni; Ghislaine J M van Thiel; Maureen Cronin; Gunnar Brobert; Panos Vardas; Stefan D Anker; Diederick E Grobbee; Spiros Denaxas Journal: Eur Heart J Date: 2018-04-21 Impact factor: 29.983
Authors: Sebastian Weichwald; Alessandro Candreva; Rebekka Burkholz; Roland Klingenberg; Lorenz Räber; Dik Heg; Robert Manka; Baris Gencer; François Mach; David Nanchen; Nicolas Rodondi; Stephan Windecker; Reijo Laaksonen; Stanley L Hazen; Arnold von Eckardstein; Frank Ruschitzka; Thomas F Lüscher; Joachim M Buhmann; Christian M Matter Journal: Eur Heart J Acute Cardiovasc Care Date: 2021-10-27
Authors: Lusha W Liang; Michael A Fifer; Kohei Hasegawa; Mathew S Maurer; Muredach P Reilly; Yuichi J Shimada Journal: Circ Genom Precis Med Date: 2021-04-23
Authors: Amitava Banerjee; Suliang Chen; Ghazaleh Fatemifar; Mohamad Zeina; R Thomas Lumbers; Johanna Mielke; Simrat Gill; Dipak Kotecha; Daniel F Freitag; Spiros Denaxas; Harry Hemingway Journal: BMC Med Date: 2021-04-06 Impact factor: 11.150