Prediction of cardiovascular outcomes with machine learning techniques: application to the Cardiovascular Outcomes in Renal Atherosclerotic Lesions (CORAL) study

Abstract

Background: Data derived from the Cardiovascular Outcomes in Renal Atherosclerotic Lesions (CORAL) study were analyzed in an effort to employ machine learning methods to predict the composite endpoint described in the original study.

Methods: We identified 573 CORAL subjects with complete baseline data and the presence or absence of a composite endpoint for the study. These data were subjected to several models including a generalized linear (logistic-linear) model, support vector machine, decision tree, feed-forward neural network, and random forest, in an effort to attempt to predict the composite endpoint. The subjects were arbitrarily divided into training and testing subsets according to an 80%:20% distribution with various seeds. Prediction models were optimized within the CARET package of R.

Results: The best performance of the different machine learning techniques was that of the random forest method which yielded a receiver operator curve (ROC) area of 68.1%±4.2% (mean ± SD) on the testing subset with ten different seed values used to separate training and testing subsets. The four most important variables in the random forest method were SBP, serum creatinine, glycosylated hemoglobin, and DBP. Each of these variables was also important in at least some of the other methods. The treatment assignment group was not consistently an important determinant in any of the models.

Conclusion: Prediction of a composite cardiovascular outcome was difficult in the CORAL population, even when employing machine learning methods. Assignment to either the stenting or best medical therapy group did not serve as an important predictor of composite outcome.

Clinical trial registration: ClinicalTrials.gov, NCT00081731.

Keywords: cardiovascular disease; chronic kidney disease; glomerular filtration rate; hypertension; ischemic renal disease; renal artery stenosis.

Conflict of interest statement

Disclosure KRT has received grants from the NHLBI and the National Institute of Diabetes, Digestive, and Kidney Diseases (NIDDK). She has received consulting fees from Eli Lilly and Company, Amgen, and Noxxon Pharma, as well as research support from Eli Lilly and Company. LDD has received grants from the National Institutes of Health (NIH) and research support from Pfizer, Astra Zeneca, and Johnson & Johnson. BAG and MS have received grants from the NIH. ST has received personal fees from AbbVie. KJ has received grants from the Medical College of Toledo. KP has received grants from the NHLBI. JMM has received personal fees from the Harvard Clinical Research Institute and grants from the NHLBI. RBD Sr has received grants from the NHLBI. DEC has received grants from the NHLBI and research support from Medtronic, Boston Scientific, and Abbott Vascular. CJC has received research funding from Cordis, study drugs from AstraZeneca, and study drugs and research funding from Pfizer. JIS has received grants from the NIH, BrickStreet Insurance Endowment, and the Huntington Foundation Endowment. The authors report no other conflicts of interest in this work.

Figures

Figure 1

Representative ROCs generated with different models with a seed of 2. Red is generalized linear, green the support vector machine, blue the decision tree, orange the neural network, and purple the random forest model. Abbreviation: ROC, receiver operator curve.