Retrospective Study
Copyright ©The Author(s) 2023. Published by Baishideng Publishing Group Inc. All rights reserved.
World J Cardiol. Mar 26, 2023; 15(3): 95-105
Published online Mar 26, 2023. doi: 10.4330/wjc.v15.i3.95
Prediction of permanent pacemaker implantation after transcatheter aortic valve replacement: The role of machine learning
Pradyumna Agasthi, Hasan Ashraf, Sai Harika Pujari, Marlene Girardo, Andrew Tseng, Farouk Mookadam, Nithin Venepally, Matthew R Buras, Bishoy Abraham, Banveet K Khetarpal, Mohamed Allam, Siva K Mulpuru MD, Mackram F Eleid, Kevin L Greason, Nirat Beohar, John Sweeney, David Fortuin, David R Jr Holmes, Reza Arsanjani
Pradyumna Agasthi, Hasan Ashraf, Farouk Mookadam, Nithin Venepally, Bishoy Abraham, Banveet K Khetarpal, Mohamed Allam, John Sweeney, David Fortuin, Reza Arsanjani, Department of Cardiology, Mayo Clinic, Phoenix, AZ 85054, United States
Sai Harika Pujari, Department of Internal Medicine, The Brooklyn Hospital Center, Brooklyn, NY 11201, United States
Marlene Girardo, Department of Biostatistics, Mayo Clinic, Phoenix, AZ 85054, United States
Andrew Tseng, Siva K Mulpuru MD, Mackram F Eleid, David R Jr Holmes, Department of Cardiovascular Diseases, Mayo Clinic, Rochester, MN 55905, United States
Matthew R Buras, Department of Statistics, Mayo Clinic, Phoenix, AZ 85054, United States
Kevin L Greason, Department of Cardiovascular Surgery, Mayo Clinic, Rochester, MN 55905, United States
Nirat Beohar, Mount Sinai Medical Center, Columbia University, Miami Beach, FL 33138, United States
Author contributions: Agasthi P conceptualizated, did methodology, project administration, review, and editing; Ashraf H methodology and writing the manuscript; Pujari SH did the data curation, visualization, and writing; Girardo M did the formal analysis, methodology, review, and editing; Tseng A writing, review, and editing the manuscript; Mookadam F did the methodology, project administration, review, and editing; Venepally N did the writing and data curation; Buras MR did the methodology and writing; Allam M did the data curation, visualization, and writing; Abraham B, Khetarpal BK, MD SKM, Eleid MF, Greason KL, Beohar N, Sweeney J, and Fortuin D writing, review and editing the manuscript; Holmes DRJ and Arsanjani R did the methodology, project administration, review, and editing.
Conflict-of-interest statement: All the authors report no relevant conflicts of interest for this article.
Data sharing statement: The authors confirm that the data supporting the findings of this study are available.
Open-Access: This article is an open-access article that was selected by an in-house editor and fully peer-reviewed by external reviewers. It is distributed in accordance with the Creative Commons Attribution NonCommercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited and the use is non-commercial. See: https://creativecommons.org/Licenses/by-nc/4.0/
Corresponding author: Sai Harika Pujari, MBBS, Department of Internal Medicine, The Brooklyn Hospital Center, 121 Dekalb Avenue, Brooklyn, NY 11201, United States. spujari@tbh.org
Received: November 25, 2022
Peer-review started: November 25, 2022
First decision: December 13, 2022
Revised: January 4, 2023
Accepted: March 1, 2023
Article in press: March 1, 2023
Published online: March 26, 2023
Processing time: 115 Days and 10.6 Hours
Abstract
BACKGROUND

Atrioventricular block requiring permanent pacemaker (PPM) implantation is an important complication of transcatheter aortic valve replacement (TAVR). Application of machine learning could potentially be used to predict pre-procedural risk for PPM.

AIM

To apply machine learning to be used to predict pre-procedural risk for PPM.

METHODS

A retrospective study of 1200 patients who underwent TAVR (January 2014-December 2017) was performed. 964 patients without prior PPM were included for a 30-d analysis and 657 patients without PPM requirement through 30 d were included for a 1-year analysis. After the exclusion of variables with near-zero variance or ≥ 50% missing data, 167 variables were included in the random forest gradient boosting algorithm (GBM) optimized using 5-fold cross-validations repeated 10 times. The receiver operator curve (ROC) for the GBM model and PPM risk score models were calculated to predict the risk of PPM at 30 d and 1 year.

RESULTS

Of 964 patients included in the 30-d analysis without prior PPM, 19.6% required PPM post-TAVR. The mean age of patients was 80.9 ± 8.7 years. 42.1 % were female. Of 657 patients included in the 1-year analysis, the mean age of the patients was 80.7 ± 8.2. Of those, 42.6% of patients were female and 26.7% required PPM at 1-year post-TAVR. The area under ROC to predict 30-d and 1-year risk of PPM for the GBM model (0.66 and 0.72) was superior to that of the PPM risk score (0.55 and 0.54) with a P value < 0.001.

CONCLUSION

The GBM model has good discrimination and calibration in identifying patients at high risk of PPM post-TAVR.

Keywords: Transcatheter aortic valve replacement, Permanent pacemaker implantation, Machine learning

Core Tip: Atrioventricular block requiring permanent pacemaker (PPM) implantation is an important complication of transcatheter aortic valve replacement. Application of machine learning could potentially be used to predict pre-procedural risk for PPM. Machine learning was used to predict patients who are at risk of developing conduction abnormalities requiring PPM at 30 d and 1 year. Our random forest machine learning model using machine learning outperforms PPM risk score model in its predictive value. Brachiocephalic to annulus distance to height ratio is the highest weighted predictor of PPM implantation at both 30-d and 1-year, which has not been previously described in the literature.