A Novel, Potentially Universal Machine Learning Algorithm to Predict Complications in Total Knee Arthroplasty

被引:17
作者
Devana, Sai K. [1 ,5 ]
Shah, Akash A. [1 ]
Lee, Changhee [2 ]
Roney, Andrew R. [1 ]
van der Schaar, Mihaela [2 ,3 ,4 ]
Soohoo, Nelson F. [1 ]
机构
[1] Univ Calif Los Angeles, Dept Orthopaed Surg, Los Angeles, CA USA
[2] Univ Calif Los Angeles, Dept Elect & Comp Engn, Los Angeles, CA USA
[3] Univ Cambridge, Dept Appl Math & Theoret Phys, Cambridge, England
[4] Alan Turing Inst, London, England
[5] 10982 Roebling Ave,Apt 337, Los Angeles, CA 90024 USA
来源
ARTHROPLASTY TODAY | 2021年 / 10卷
关键词
AutoPrognosis; Machine learning; Predictive modeling; Knee replacement; PERIPROSTHETIC JOINT INFECTION; TOTAL HIP-ARTHROPLASTY; RISK CALCULATOR; 30-DAY MORTALITY; OUTCOMES; ASSOCIATION; REPLACEMENT; VETERANS; VOLUME;
D O I
10.1016/j.artd.2021.06.020
中图分类号
R826.8 [整形外科学]; R782.2 [口腔颌面部整形外科学]; R726.2 [小儿整形外科学]; R62 [整形外科学(修复外科学)];
学科分类号
摘要
Background: There remains a lack of accurate and validated outcome-prediction models in total knee arthroplasty (TKA). While machine learning (ML) is a powerful predictive tool, determining the proper algorithm to apply across diverse data sets is challenging. AutoPrognosis (AP) is a novel method that uses automated ML framework to incorporate the best performing stages of prognostic modeling into a single well-calibrated algorithm. We aimed to compare various ML methods to AP in predictive performance of complications after TKA. Methods: Thirty-eight preoperative patient demographics and clinical features from all primary TKAs performed at California-licensed hospitals between 2015 and 2017 were evaluated as predictors of major complications after TKA. Traditional logistic regression (LR), various other ML methods (XGBoost, Gradient Boosting, AdaBoost, and Random Forest), and AP were used for model building to determine discriminative power (area under receiver operating curve), calibration (Brier score), and feature importance. Results: Between 2015 and 2017, there were a total of 156,750 TKAs with 1109 (0.7%) total major complications. AP had the highest discriminative performance with area under receiver operating curve 0.679 compared with LR, XGBoost, Gradient Boosting, AdaBoost, and Random Forest (0.617, 0.601, 0.662, 0.657, and 0.545, respectively). AP (Brier score 0.007) had similar calibration as the other ML methods (0.006, 0.006, 0.022, 0.007, and 0.008, respectively). The variables that are most important for AP differ from those that are most important for LR. Conclusion: Compared to conventional ML algorithms, AP has superior discriminative ability with similar calibration and suggests nonlinear relationships between variables in outcomes of TKA. (c) 2021 The Authors. Published by Elsevier Inc. on behalf of The American Association of Hip and Knee Surgeons. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/ licenses/by-nc-nd/4.0/).
引用
收藏
页码:135 / 143
页数:9
相关论文
共 49 条
[1]  
Alaa AM, 2018, PR MACH LEARN RES, V80
[2]   Cardiovascular disease risk prediction using automated machine learning: A prospective study of 423,604 UK Biobank participants [J].
Alaa, Ahmed M. ;
Bolton, Thomas ;
Di Angelantonio, Emanuele ;
Rudd, James H. F. ;
van der Schaar, Mihaela .
PLOS ONE, 2019, 14 (05)
[3]   Prognostication and Risk Factors for Cystic Fibrosis via Automated Machine Learning [J].
Alaa, Ahmed M. ;
van der Schaar, Mihaela .
SCIENTIFIC REPORTS, 2018, 8
[4]   Thirty-Day Postoperative Complications and Mortality Following Total Knee Arthroplasty [J].
Belmont, Philip J., Jr. ;
Goodman, Gens P. ;
Waterman, Brian R. ;
Bader, Julia O. ;
Schoenfeld, Andrew J. .
JOURNAL OF BONE AND JOINT SURGERY-AMERICAN VOLUME, 2014, 96A (01) :20-26
[5]   Development and Evaluation of the Universal ACS NSQIP Surgical Risk Calculator: A Decision Aid and Informed Consent Tool for Patients and Surgeons [J].
Bilimoria, Karl Y. ;
Liu, Yaoming ;
Paruch, Jennifer L. ;
Zhou, Lynn ;
Kmiecik, Thomas E. ;
Ko, Clifford Y. ;
Cohen, Mark E. .
JOURNAL OF THE AMERICAN COLLEGE OF SURGEONS, 2013, 217 (05) :833-+
[6]   Infection after total knee arthroplasty [J].
Blom, AW ;
Brown, J ;
Taylor, AH ;
Pattison, G ;
Whitehouse, S ;
Bannister, GC .
JOURNAL OF BONE AND JOINT SURGERY-BRITISH VOLUME, 2004, 86B (05) :688-691
[7]   Hypoalbuminemia Independently Predicts Surgical Site Infection, Pneumonia, Length of Stay, and Readmission After Total Joint Arthroplasty [J].
Bohl, Daniel D. ;
Shen, Mary R. ;
Kayupov, Erdan ;
Della Valle, Craig J. .
JOURNAL OF ARTHROPLASTY, 2016, 31 (01) :15-21
[8]   Estimating Risk in Medicare Patients With THA: An Electronic Risk Calculator for Periprosthetic Joint Infection and Mortality [J].
Bozic, Kevin J. ;
Ong, Kevin ;
Lau, Edmund ;
Berry, Daniel J. ;
Vail, Thomas P. ;
Kurtz, Steven M. ;
Rubash, Harry E. .
CLINICAL ORTHOPAEDICS AND RELATED RESEARCH, 2013, 471 (02) :574-583
[9]   Patient-Related Risk Factors for Periprosthetic Joint Infection and Postoperative Mortality Following Total Hip Arthroplasty in Medicare Patients [J].
Bozic, Kevin J. ;
Lau, Edmund ;
Kurtz, Steven ;
Ong, Kevin ;
Rubash, Harry ;
Vail, Thomas P. ;
Berry, Daniel J. .
JOURNAL OF BONE AND JOINT SURGERY-AMERICAN VOLUME, 2012, 94A (09) :794-800
[10]   The impact of infection after total hip arthroplasty on hospital and surgeon resource utilization [J].
Bozic, KJ ;
Ries, MD .
JOURNAL OF BONE AND JOINT SURGERY-AMERICAN VOLUME, 2005, 87A (08) :1746-1751