TY - JOUR

T1 - Early prediction of clinical deterioration using data-driven machine-learning modeling of electronic health records

AU - Ruiz, Victor M.

AU - Goldsmith, Michael P.

AU - Shi, Lingyun

AU - Simpao, Allan F.

AU - Gálvez, Jorge A.

AU - Naim, Maryam Y.

AU - Nadkarni, Vinay

AU - Gaynor, J. William

AU - Tsui, Fuchiang (Rich)

N1 - Funding Information: This work was supported by Children's Hospital of Philadelphia ( CHOP ) and the Richard King Mellon Foundation (MWRIF 3659). The content is solely the responsibility of the authors and does not necessarily represent the official views of CHOP or of the Richard King Mellon Foundation. Michael Goldsmith and Vinay Nadkarni have National Institutes of Health / National Heart, Lung, and Blood Institute Small Business Innovation Research Program grant funding to their institution to investigate risk assessments created by Etiometry, Inc, in postoperative patients (2R44HL117340-03A1/04/05). Publisher Copyright: © 2021 The American Association for Thoracic Surgery

PY - 2022

Y1 - 2022

N2 - Objectives: To develop and evaluate a high-dimensional, data-driven model to identify patients at high risk of clinical deterioration from routinely collected electronic health record (EHR) data. Materials and Methods: In this single-center, retrospective cohort study, 488 patients with single-ventricle and shunt-dependent congenital heart disease <6 months old were admitted to the cardiac intensive care unit before stage 2 palliation between 2014 and 2019. Using machine-learning techniques, we developed the Intensive care Warning Index (I-WIN), which systematically assessed 1028 regularly collected EHR variables (vital signs, medications, laboratory tests, and diagnoses) to identify patients in the cardiac intensive care unit at elevated risk of clinical deterioration. An ensemble of 5 extreme gradient boosting models was developed and validated on 203 cases (130 emergent endotracheal intubations, 34 cardiac arrests requiring cardiopulmonary resuscitation, 10 extracorporeal membrane oxygenation cannulations, and 29 cardiac arrests requiring cardiopulmonary resuscitation onto extracorporeal membrane oxygenation) and 378 control periods from 446 patients. Results: At 4 hours before deterioration, the model achieved an area under the receiver operating characteristic curve of 0.92 (95% confidence interval, 0.84-0.98), 0.881 sensitivity, 0.776 positive predictive value, 0.862 specificity, and 0.571 Brier skill score. Performance remained high at 8 hours before deterioration with 0.815 (0.688-0.921) area under the receiver operating characteristic curve. Conclusions: I-WIN accurately predicted deterioration events in critically-ill infants with high-risk congenital heart disease up to 8 hours before deterioration, potentially allowing clinicians to target interventions. We propose a paradigm shift from conventional expert consensus–based selection of risk factors to a data-driven, machine-learning methodology for risk prediction. With the increased availability of data capture in EHRs, I-WIN can be extended to broader applications in data-rich environments in critical care.

AB - Objectives: To develop and evaluate a high-dimensional, data-driven model to identify patients at high risk of clinical deterioration from routinely collected electronic health record (EHR) data. Materials and Methods: In this single-center, retrospective cohort study, 488 patients with single-ventricle and shunt-dependent congenital heart disease <6 months old were admitted to the cardiac intensive care unit before stage 2 palliation between 2014 and 2019. Using machine-learning techniques, we developed the Intensive care Warning Index (I-WIN), which systematically assessed 1028 regularly collected EHR variables (vital signs, medications, laboratory tests, and diagnoses) to identify patients in the cardiac intensive care unit at elevated risk of clinical deterioration. An ensemble of 5 extreme gradient boosting models was developed and validated on 203 cases (130 emergent endotracheal intubations, 34 cardiac arrests requiring cardiopulmonary resuscitation, 10 extracorporeal membrane oxygenation cannulations, and 29 cardiac arrests requiring cardiopulmonary resuscitation onto extracorporeal membrane oxygenation) and 378 control periods from 446 patients. Results: At 4 hours before deterioration, the model achieved an area under the receiver operating characteristic curve of 0.92 (95% confidence interval, 0.84-0.98), 0.881 sensitivity, 0.776 positive predictive value, 0.862 specificity, and 0.571 Brier skill score. Performance remained high at 8 hours before deterioration with 0.815 (0.688-0.921) area under the receiver operating characteristic curve. Conclusions: I-WIN accurately predicted deterioration events in critically-ill infants with high-risk congenital heart disease up to 8 hours before deterioration, potentially allowing clinicians to target interventions. We propose a paradigm shift from conventional expert consensus–based selection of risk factors to a data-driven, machine-learning methodology for risk prediction. With the increased availability of data capture in EHRs, I-WIN can be extended to broader applications in data-rich environments in critical care.

KW - cardiopulmonary resuscitation

KW - electronic health records

KW - extracorporeal membrane oxygenation

KW - intratracheal

KW - intubation

KW - machine learning

KW - univentricular heart

U2 - 10.1016/j.jtcvs.2021.10.060

DO - 10.1016/j.jtcvs.2021.10.060

M3 - Journal article

C2 - 34949457

AN - SCOPUS:85121013286

VL - 164

SP - 211-222.e3

JO - Journal of Thoracic and Cardiovascular Surgery

JF - Journal of Thoracic and Cardiovascular Surgery

SN - 0022-5223

IS - 1

ER -