P Elliott Miller1, Sumeet Pawar1, Benjamin Vaccaro1, Megan McCullough1, Pooja Rao2, Rohit Ghosh2, Prashant Warier2, Nihar R Desai3, Tariq Ahmad4. 1. Section of Cardiovascular Medicine, Yale School of Medicine, New Haven, Connecticut. 2. Qure.ai, Mumbai, India. 3. Section of Cardiovascular Medicine, Yale School of Medicine, New Haven, Connecticut; Center for Outcomes Research and Evaluation, New Haven, Connecticut. 4. Section of Cardiovascular Medicine, Yale School of Medicine, New Haven, Connecticut. Electronic address: tariq.ahmad@yale.edu.
Abstract
BACKGROUND: Traditional statistical approaches to prediction of outcomes have drawbacks when applied to large clinical databases. It is hypothesized that machine learning methodologies might overcome these limitations by considering higher-dimensional and nonlinear relationships among patient variables. METHODS AND RESULTS: The Unified Network for Organ Sharing (UNOS) database was queried from 1987 to 2014 for adult patients undergoing cardiac transplantation. The dataset was divided into 3 time periods corresponding to major allocation adjustments and based on geographic regions. For our outcome of 1-year survival, we used the standard statistical methods logistic regression, ridge regression, and regressions with LASSO (least absolute shrinkage and selection operator) and compared them with the machine learning methodologies neural networks, naïve-Bayes, tree-augmented naïve-Bayes, support vector machines, random forest, and stochastic gradient boosting. Receiver operating characteristic curves and C-statistics were calculated for each model. C-Statistics were used for comparison of discriminatory capacity across models in the validation sample. After identifying 56,477 patients, the major univariate predictors of 1-year survival after heart transplantation were consistent with earlier reports and included age, renal function, body mass index, liver function tests, and hemodynamics. Advanced analytic models demonstrated similarly modest discrimination capabilities compared with traditional models (C-statistic ≤0.66, all). The neural network model demonstrated the highest C-statistic (0.66) but this was only slightly superior to the simple logistic regression, ridge regression, and regression with LASSO models (C-statistic = 0.65, all). Discrimination did not vary significantly across the 3 historically important time periods. CONCLUSIONS: The use of advanced analytic algorithms did not improve prediction of 1-year survival from heart transplant compared with more traditional prediction models. The prognostic abilities of machine learning techniques may be limited by quality of the clinical dataset.
BACKGROUND: Traditional statistical approaches to prediction of outcomes have drawbacks when applied to large clinical databases. It is hypothesized that machine learning methodologies might overcome these limitations by considering higher-dimensional and nonlinear relationships among patient variables. METHODS AND RESULTS: The Unified Network for Organ Sharing (UNOS) database was queried from 1987 to 2014 for adult patients undergoing cardiac transplantation. The dataset was divided into 3 time periods corresponding to major allocation adjustments and based on geographic regions. For our outcome of 1-year survival, we used the standard statistical methods logistic regression, ridge regression, and regressions with LASSO (least absolute shrinkage and selection operator) and compared them with the machine learning methodologies neural networks, naïve-Bayes, tree-augmented naïve-Bayes, support vector machines, random forest, and stochastic gradient boosting. Receiver operating characteristic curves and C-statistics were calculated for each model. C-Statistics were used for comparison of discriminatory capacity across models in the validation sample. After identifying 56,477 patients, the major univariate predictors of 1-year survival after heart transplantation were consistent with earlier reports and included age, renal function, body mass index, liver function tests, and hemodynamics. Advanced analytic models demonstrated similarly modest discrimination capabilities compared with traditional models (C-statistic ≤0.66, all). The neural network model demonstrated the highest C-statistic (0.66) but this was only slightly superior to the simple logistic regression, ridge regression, and regression with LASSO models (C-statistic = 0.65, all). Discrimination did not vary significantly across the 3 historically important time periods. CONCLUSIONS: The use of advanced analytic algorithms did not improve prediction of 1-year survival from heart transplant compared with more traditional prediction models. The prognostic abilities of machine learning techniques may be limited by quality of the clinical dataset.
Authors: Leah L Zullig; Shelley A Jazowski; Tracy Y Wang; Anne Hellkamp; Daniel Wojdyla; Laine Thomas; Lisa Egbuonu-Davis; Anne Beal; Hayden B Bosworth Journal: Health Serv Res Date: 2019-08-20 Impact factor: 3.402
Authors: Thomas G Myers; Prem N Ramkumar; Benjamin F Ricciardi; Kenneth L Urish; Jens Kipper; Constantinos Ketonis Journal: J Bone Joint Surg Am Date: 2020-05-06 Impact factor: 5.284
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
Authors: Santino R Rellum; Jaap Schuurmans; Ward H van der Ven; Susanne Eberl; Antoine H G Driessen; Alexander P J Vlaar; Denise P Veelo Journal: J Thorac Dis Date: 2021-12 Impact factor: 2.895
Authors: Neta Gotlieb; Amirhossein Azhie; Divya Sharma; Ashley Spann; Nan-Ji Suo; Jason Tran; Ani Orchanian-Cheff; Bo Wang; Anna Goldenberg; Michael Chassé; Heloise Cardinal; Joseph Paul Cohen; Andrea Lodi; Melanie Dieude; Mamatha Bhat Journal: NPJ Digit Med Date: 2022-07-11
Authors: Jeremy A Balch; Daniel Delitto; Patrick J Tighe; Ali Zarrinpar; Philip A Efron; Parisa Rashidi; Gilbert R Upchurch; Azra Bihorac; Tyler J Loftus Journal: Front Immunol Date: 2021-09-16 Impact factor: 7.561