Advertisement

Evaluation of Random Forest and Ensemble Methods at Predicting Complications Following Cardiac Surgery

  • Linda LappEmail author
  • Matt-Mouley Bouamrane
  • Kimberley Kavanagh
  • Marc Roper
  • David Young
  • Stefan Schraag
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 11526)

Abstract

Cardiac patients undergoing surgery face increased risk of postoperative complications, due to a combination of factors, including higher risk surgery, their age at time of surgery and the presence of co-morbid conditions. They will therefore require high levels of care and clinical resources throughout their perioperative journey (i.e. before, during and after surgery). Although surgical mortality rates in the UK have remained low, postoperative complications on the other hand are common and can have a significant impact on patients’ quality of life, increase hospital length of stay and healthcare costs. In this study we used and compared several machine learning methods – random forest, AdaBoost, gradient boosting model and stacking – to predict severe postoperative complications after cardiac surgery based on preoperative variables obtained from a surgical database of a large acute care hospital in Scotland. Our results show that AdaBoost has the best overall performance (AUC = 0.731), and also outperforms EuroSCORE and EuroSCORE II in other studies predicting postoperative complications. Random forest (Sensitivity = 0.852, negative predictive value = 0.923), however, and gradient boosting model (Sensitivity = 0.875 and negative predictive value = 0.920) have the best performance at predicting severe postoperative complications based on sensitivity and negative predictive value.

Keywords

Postoperative complications Machine learning Cardiac surgery 

References

  1. 1.
    Findlay, G.P., Goodwin, A.P.L., Protopapa, K., Smith, N.C.E., Mason, M.: Knowing the risk: a review of the peri-operative care of surgical patients, London (2011)Google Scholar
  2. 2.
    Bouamrane, M.-M., Mair, F.S.: Implementation of an integrated preoperative care pathway and regional electronic clinical portal for preoperative assessment. BMC Med. Inform. Decis. Mak. 14, 93 (2014)CrossRefGoogle Scholar
  3. 3.
    Bouamrane, M.-M., Mair, F.S.: A study of clinical and information management processes in the surgical pre-assessment clinic. BMC Med. Inform. Decis. Mak. 14, 22 (2014)CrossRefGoogle Scholar
  4. 4.
    Moonesinghe, S.R., Mythen, M.G., Das, P., Rowan, K.M., Grocott, M.P.: Risk stratification tools for predicting morbidity and mortality in adult patients undergoing major surgery: qualitative systematic review. Anesthesiology 119, 959–981 (2013)CrossRefGoogle Scholar
  5. 5.
    SCTS: Blue Book Online. http://bluebook.scts.org/
  6. 6.
    Pinto, A., Faiz, O., Davis, R., Almoudaris, A., Vincent, C.: Surgical complications and their impact on patients’ psychosocial well-being: a systematic review and meta-analysis. BMJ Open 6, e007224 (2016)CrossRefGoogle Scholar
  7. 7.
    Knapik, P., Ciesla, D., Borowik, D., Czempik, P., Knapik, T.: Prolonged ventilation post cardiac surgery - tips and pitfalls of the prediction game. J. Cardiothorac. Surg. 6, 158 (2011)CrossRefGoogle Scholar
  8. 8.
    Eappen, S., et al.: Relationship between occurrence of surgical complications and hospital finances. JAMA 309, 1599–1606 (2013)CrossRefGoogle Scholar
  9. 9.
    Roques, F., Michel, P., Goldstone, A.R., Nashef, S.A.M.: The logistic EuroSCORE. Eur. Heart J. 24, 1–2 (2003)CrossRefGoogle Scholar
  10. 10.
    Nashef, S.A., et al.: EuroSCORE II. Eur. J. Cardio-Thorac. Surg. 41, 734–744 (2012)CrossRefGoogle Scholar
  11. 11.
    Barnett, S., Moonesinghe, S.R.: Clinical risk scores to guide perioperative management. Postgrad. Med. J. 87, 535–541 (2011)CrossRefGoogle Scholar
  12. 12.
    Chawla, N.V., Bowyer, K.W., Hall, L.O., Kegelmeyer, W.P.: SMOTE: synthetic minority oversampling technique. J. Artif. Intell. Res. 16, 321–357 (2002)CrossRefGoogle Scholar
  13. 13.
    Galar, M., Fernandez, A., Barrenechea, E., Bustince, H., Herrera, F.: Review on Ensembles for the Class Imbalance Problem: Bagging-, Boosting-, and Hybrid-Based Approaches. IEEE Trans. Syst. Man Cybern. Part C Appl. Rev. 42, 463–484 (2012)CrossRefGoogle Scholar
  14. 14.
    Kuncheva, L.I.: Combining Patterns Classifiers: Methods and Algorithms. Wiley, New York (2004)CrossRefGoogle Scholar
  15. 15.
    Bischl, B., Mersmann, O., Trautmann, H., Weihs, C.: Resampling methods for meta-model validation with recommendations for evolutionary computation. Evol. Comput. 20, 249–275 (2012)CrossRefGoogle Scholar
  16. 16.
    Freund, Y., Schapire, R.E.: Experiments with a new boosting algorithm. In: The Thirteenth International Conference on Machine Learning, pp. 148–156 (1996)Google Scholar
  17. 17.
    Friedman, J.H.: Greedy function approximation: a gradient boosting machine. Ann. Stat. 29, 1189–1232 (2001)MathSciNetCrossRefGoogle Scholar
  18. 18.
    Walker, S.H., Duncan, D.B.: Estimation of the probability of an event as a function of several independent variables. Biometrika 54, 167–179 (1967)MathSciNetCrossRefGoogle Scholar
  19. 19.
    Ho, T.K.: The random subspace method for constructing decision forests. IEEE Trans. Pattern Anal. Mach. Intell. 20, 832–844 (1998)CrossRefGoogle Scholar
  20. 20.
    Zhang, H.: The optimality of naive Bayes. In: FLAIRS2004 Conference (2004)Google Scholar
  21. 21.
    Breiman, L.: Bagging predictors. Mach. Learn. 24, 123–140 (1996)zbMATHGoogle Scholar
  22. 22.
    Valverde-Albacete, F.J., Pelaez-Moreno, C.: 100% classification accuracy considered harmful: the normalized information transfer factor explains the accuracy paradox. PLoS ONE 9, e84217 (2014)CrossRefGoogle Scholar
  23. 23.
    Wang, T.K., Li, A.Y., Ramanathan, T., Stewart, R.A., Gamble, G., White, H.D.: Comparison of four risk scores for contemporary isolated coronary artery bypass grafting. Hear. Lung Circ. 23, 469–474 (2014)CrossRefGoogle Scholar
  24. 24.
    Reis, C., Barbiero, S.M., Ribas, L.: The effect of the body mass index on postoperative complications of coronary artery bypass grafting in elderly. Braz. J. Cardiovasc. Surg. 23, 524–529 (2008)CrossRefGoogle Scholar
  25. 25.
    Kaul, P., Naylor, C.D., Armstrong, P.W., Mark, D.B., Theroux, P., Dagenais, G.R.: Assessment of activity status and survival according to the Canadian Cardiovascular Society angina classification. Can. J. Cardiol. 25, e225–e231 (2009)CrossRefGoogle Scholar
  26. 26.
    Ji, Q., Zhao, H., Mei, Y., Shi, Y., Ma, R., Ding, W.: Impact of smoking on early clinical outcomes in patients undergoing coronary artery bypass grafting surgery. J. Cardiothorac. Surg. 10, 16 (2015)CrossRefGoogle Scholar
  27. 27.
    Sharif-Kashani, B., et al.: Smoking and wound complications after coronary artery bypass grafting. J. Surg. Res. 200, 743–748 (2016)CrossRefGoogle Scholar
  28. 28.
    Geissler, H.J., et al.: Risk stratification in heart surgery: comparison of six score systems. Eur. J. Cardio-Thorac. Surg. 17, 400–406 (2000)CrossRefGoogle Scholar
  29. 29.
    Hirose, H., et al.: EuroSCORE predicts postoperative mortality, certain morbidities, and recovery time. Interact. Cardiovasc. Thorac. Surg. 9, 613–617 (2009)CrossRefGoogle Scholar
  30. 30.
    Pitkänen, O., Niskanen, M., Rehnberg, S., Hippelainen, M., Hynynen, M.: Intra-institutional prediction of outcome after cardiac surgery: comparison between a locally derived model and the EuroSCORE. Eur. J. Cardiothorac. Surg. 18, 703–710 (2000)CrossRefGoogle Scholar
  31. 31.
    Wang, T.K.M., Harmos, S., Gamble, G.D., Ramanathan, T., Ruygrok, P.N.: Performance of contemporary surgical risk scores for mitral valve surgery. J. Card. Surg. 32, 172–176 (2017)CrossRefGoogle Scholar
  32. 32.
    Lutkenhoner, B., Basel, T.: Predictive modeling for diagnostic tests with high specificity, but low sensitivity: a study of the glycerol test in patients with suspected meniere’s disease. PLoS ONE 8, e79315 (2013)CrossRefGoogle Scholar
  33. 33.
    McEvoy, J.W., et al.: Risk and the physics of clinical prediction. Am. J. Cardiol. 113, 1429–1435 (2014)CrossRefGoogle Scholar

Copyright information

© Springer Nature Switzerland AG 2019

Authors and Affiliations

  1. 1.University of StrathclydeGlasgowUK
  2. 2.Golden Jubilee National HospitalClydebankUK

Personalised recommendations