Utilizing electronic health data and machine learning for the prediction of 30-day unplanned readmission or all-cause mortality in heart failure
Background: Existing risk assessment tools for heart failure (HF) outcomes use structured databases with static, single-timepoint clinical data and have limited accuracy. Objective: The purpose of this study was to develop a comprehensive approach for accurate prediction of 30-day unplanned readmiss...
Main Authors: | , , , , , , , , , , , , , , |
---|---|
Format: | Article |
Language: | English |
Published: |
Elsevier
2020-09-01
|
Series: | Cardiovascular Digital Health Journal |
Subjects: | |
Online Access: | http://www.sciencedirect.com/science/article/pii/S2666693620300104 |
id |
doaj-ba18048440d746b598b2dae86aead55a |
---|---|
record_format |
Article |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Ashley N. Beecy, MD Manasa Gummalla, BA Evan Sholle, MS Zhuoran Xu, MSc Yiye Zhang, MSc, PhD Kelly Michalak, BA Kristina Dolan, BA Yasin Hussain, MD Benjamin C. Lee, PhD Yongkang Zhang, PhD Parag Goyal, MSc, MD Thomas R. Campion, Jr., PhD Leslee J. Shaw, PhD Lohendran Baskaran, MBBS Subhi J. Al’Aref, MD |
spellingShingle |
Ashley N. Beecy, MD Manasa Gummalla, BA Evan Sholle, MS Zhuoran Xu, MSc Yiye Zhang, MSc, PhD Kelly Michalak, BA Kristina Dolan, BA Yasin Hussain, MD Benjamin C. Lee, PhD Yongkang Zhang, PhD Parag Goyal, MSc, MD Thomas R. Campion, Jr., PhD Leslee J. Shaw, PhD Lohendran Baskaran, MBBS Subhi J. Al’Aref, MD Utilizing electronic health data and machine learning for the prediction of 30-day unplanned readmission or all-cause mortality in heart failure Cardiovascular Digital Health Journal Big data Electronic health data Heart failure Machine learning Readmission |
author_facet |
Ashley N. Beecy, MD Manasa Gummalla, BA Evan Sholle, MS Zhuoran Xu, MSc Yiye Zhang, MSc, PhD Kelly Michalak, BA Kristina Dolan, BA Yasin Hussain, MD Benjamin C. Lee, PhD Yongkang Zhang, PhD Parag Goyal, MSc, MD Thomas R. Campion, Jr., PhD Leslee J. Shaw, PhD Lohendran Baskaran, MBBS Subhi J. Al’Aref, MD |
author_sort |
Ashley N. Beecy, MD |
title |
Utilizing electronic health data and machine learning for the prediction of 30-day unplanned readmission or all-cause mortality in heart failure |
title_short |
Utilizing electronic health data and machine learning for the prediction of 30-day unplanned readmission or all-cause mortality in heart failure |
title_full |
Utilizing electronic health data and machine learning for the prediction of 30-day unplanned readmission or all-cause mortality in heart failure |
title_fullStr |
Utilizing electronic health data and machine learning for the prediction of 30-day unplanned readmission or all-cause mortality in heart failure |
title_full_unstemmed |
Utilizing electronic health data and machine learning for the prediction of 30-day unplanned readmission or all-cause mortality in heart failure |
title_sort |
utilizing electronic health data and machine learning for the prediction of 30-day unplanned readmission or all-cause mortality in heart failure |
publisher |
Elsevier |
series |
Cardiovascular Digital Health Journal |
issn |
2666-6936 |
publishDate |
2020-09-01 |
description |
Background: Existing risk assessment tools for heart failure (HF) outcomes use structured databases with static, single-timepoint clinical data and have limited accuracy. Objective: The purpose of this study was to develop a comprehensive approach for accurate prediction of 30-day unplanned readmission and all-cause mortality (ACM) that integrates clinical and physiological data available in the electronic health record system. Methods: Three predictive models for 30-day unplanned readmissions or ACM were created using an extreme gradient boosting approach: (1) index admission model; (2) index discharge model; and (3) feature-aggregated model. Performance was assessed by the area under the curve (AUC) metric and compared with that of the HOSPITAL score, a widely used predictive model for hospital readmission. Results: A total of 3774 patients with a primary billing diagnosis of HF were included (614 experienced the primary outcome), with 796 variables used in the admission and discharge models, and 2032 in the feature-aggregated model. The index admission model had AUC = 0.723, the index discharge model had AUC = 0.754, and the feature-aggregated model had AUC = 0.756 for prediction of 30-day unplanned readmission or ACM. For comparison, the HOSPITAL score had AUC = 0.666 (admission model: P = .093; discharge model: P = .022; feature aggregated: P = .012). Conclusion: These models predict risk of HF hospitalizations and ACM in patients admitted with HF and emphasize the importance of incorporating large numbers of variables in machine learning models to identify predictors for future investigation. |
topic |
Big data Electronic health data Heart failure Machine learning Readmission |
url |
http://www.sciencedirect.com/science/article/pii/S2666693620300104 |
work_keys_str_mv |
AT ashleynbeecymd utilizingelectronichealthdataandmachinelearningforthepredictionof30dayunplannedreadmissionorallcausemortalityinheartfailure AT manasagummallaba utilizingelectronichealthdataandmachinelearningforthepredictionof30dayunplannedreadmissionorallcausemortalityinheartfailure AT evanshollems utilizingelectronichealthdataandmachinelearningforthepredictionof30dayunplannedreadmissionorallcausemortalityinheartfailure AT zhuoranxumsc utilizingelectronichealthdataandmachinelearningforthepredictionof30dayunplannedreadmissionorallcausemortalityinheartfailure AT yiyezhangmscphd utilizingelectronichealthdataandmachinelearningforthepredictionof30dayunplannedreadmissionorallcausemortalityinheartfailure AT kellymichalakba utilizingelectronichealthdataandmachinelearningforthepredictionof30dayunplannedreadmissionorallcausemortalityinheartfailure AT kristinadolanba utilizingelectronichealthdataandmachinelearningforthepredictionof30dayunplannedreadmissionorallcausemortalityinheartfailure AT yasinhussainmd utilizingelectronichealthdataandmachinelearningforthepredictionof30dayunplannedreadmissionorallcausemortalityinheartfailure AT benjamincleephd utilizingelectronichealthdataandmachinelearningforthepredictionof30dayunplannedreadmissionorallcausemortalityinheartfailure AT yongkangzhangphd utilizingelectronichealthdataandmachinelearningforthepredictionof30dayunplannedreadmissionorallcausemortalityinheartfailure AT paraggoyalmscmd utilizingelectronichealthdataandmachinelearningforthepredictionof30dayunplannedreadmissionorallcausemortalityinheartfailure AT thomasrcampionjrphd utilizingelectronichealthdataandmachinelearningforthepredictionof30dayunplannedreadmissionorallcausemortalityinheartfailure AT lesleejshawphd utilizingelectronichealthdataandmachinelearningforthepredictionof30dayunplannedreadmissionorallcausemortalityinheartfailure AT lohendranbaskaranmbbs utilizingelectronichealthdataandmachinelearningforthepredictionof30dayunplannedreadmissionorallcausemortalityinheartfailure AT subhijalarefmd utilizingelectronichealthdataandmachinelearningforthepredictionof30dayunplannedreadmissionorallcausemortalityinheartfailure |
_version_ |
1721390193341104128 |
spelling |
doaj-ba18048440d746b598b2dae86aead55a2021-06-08T04:44:06ZengElsevierCardiovascular Digital Health Journal2666-69362020-09-01127179Utilizing electronic health data and machine learning for the prediction of 30-day unplanned readmission or all-cause mortality in heart failureAshley N. Beecy, MD0Manasa Gummalla, BA1Evan Sholle, MS2Zhuoran Xu, MSc3Yiye Zhang, MSc, PhD4Kelly Michalak, BA5Kristina Dolan, BA6Yasin Hussain, MD7Benjamin C. Lee, PhD8Yongkang Zhang, PhD9Parag Goyal, MSc, MD10Thomas R. Campion, Jr., PhD11Leslee J. Shaw, PhD12Lohendran Baskaran, MBBS13Subhi J. Al’Aref, MD14Division of Cardiology, Department of Medicine, Weill Cornell Medicine, New York, New York; Address reprint requests and correspondence: Dr Ashley N. Beecy, Weill Cornell Medical College, 520 E 70th St, Starr 4, New York, NY 10021.Dalio Institute of Cardiovascular Imaging, Weill Cornell Medicine, New York, New YorkInformation Technologies & Services, Weill Cornell Medicine, New York, New YorkDalio Institute of Cardiovascular Imaging, Weill Cornell Medicine, New York, New York; Department of Radiology, New York-Presbyterian Hospital and Weill Cornell Medicine, New York, New YorkDepartment of Population Health Sciences, Weill Cornell Medicine, New York, New YorkDalio Institute of Cardiovascular Imaging, Weill Cornell Medicine, New York, New YorkDalio Institute of Cardiovascular Imaging, Weill Cornell Medicine, New York, New YorkYale Cardiovascular Medicine, New Haven, ConnecticutDalio Institute of Cardiovascular Imaging, Weill Cornell Medicine, New York, New YorkDepartment of Population Health Sciences, Weill Cornell Medicine, New York, New YorkDivision of Cardiology, Department of Medicine, Weill Cornell Medicine, New York, New YorkInformation Technologies & Services, Weill Cornell Medicine, New York, New York; Department of Population Health Sciences, Weill Cornell Medicine, New York, New YorkDalio Institute of Cardiovascular Imaging, Weill Cornell Medicine, New York, New York; Department of Radiology, New York-Presbyterian Hospital and Weill Cornell Medicine, New York, New YorkDalio Institute of Cardiovascular Imaging, Weill Cornell Medicine, New York, New York; Department of Radiology, New York-Presbyterian Hospital and Weill Cornell Medicine, New York, New York; Department of Cardiovascular Medicine, National Heart Centre, SingaporeDivision of Cardiology, Department of Medicine, University of Arkansas for Medical Sciences, Little Rock, ArkansasBackground: Existing risk assessment tools for heart failure (HF) outcomes use structured databases with static, single-timepoint clinical data and have limited accuracy. Objective: The purpose of this study was to develop a comprehensive approach for accurate prediction of 30-day unplanned readmission and all-cause mortality (ACM) that integrates clinical and physiological data available in the electronic health record system. Methods: Three predictive models for 30-day unplanned readmissions or ACM were created using an extreme gradient boosting approach: (1) index admission model; (2) index discharge model; and (3) feature-aggregated model. Performance was assessed by the area under the curve (AUC) metric and compared with that of the HOSPITAL score, a widely used predictive model for hospital readmission. Results: A total of 3774 patients with a primary billing diagnosis of HF were included (614 experienced the primary outcome), with 796 variables used in the admission and discharge models, and 2032 in the feature-aggregated model. The index admission model had AUC = 0.723, the index discharge model had AUC = 0.754, and the feature-aggregated model had AUC = 0.756 for prediction of 30-day unplanned readmission or ACM. For comparison, the HOSPITAL score had AUC = 0.666 (admission model: P = .093; discharge model: P = .022; feature aggregated: P = .012). Conclusion: These models predict risk of HF hospitalizations and ACM in patients admitted with HF and emphasize the importance of incorporating large numbers of variables in machine learning models to identify predictors for future investigation.http://www.sciencedirect.com/science/article/pii/S2666693620300104Big dataElectronic health dataHeart failureMachine learningReadmission |