Applying Deep Learning in Heart Failure: Hospital Readmission is Not Like Other Health Quality Metrics
Preprint
Abstract
Background Early identification of heart failure patients at increased risk for near-term adverse outcomes would assist clinicians in efficient resource allocation and improved care. Deep learning can improve identification of these patients. Methods This retrospective study examined adult heart failure patients admitted to a tertiary care institution between January 2009 and December 2018. A deep learning model was constructed with a dense input layer, three long short-term memory (LSTM) layers, and a dense hidden layer to cohesively extract features from time-series and non-time-series EHR data. Primary outcomes were all-cause hospital readmission or death within 30 days after hospital discharge. Results Among a final subset of 49,675 heart failure patients, we identified 171,563 hospital admissions described by 330 million EHR data points. There were 22,111 (13%) admissions followed by adverse 30-day outcomes, including 19,122 readmissions (87%) and mortality in 3,330 patients (15%). Our final deep learning model achieved an area under the receiver-operator characteristic curve (AUC) of 0.613 and precision-recall (PR) AUC of 0.38. Conclusions This EHR-based deep learning model developed from a decade of heart failure care achieved marginal clinical accuracy in predicting very early hospital readmission or death despite previous accurate prediction of 1-year mortality in this large study cohort. These findings suggest that factors unavailable in standard EHR data play pivotal roles in influencing early hospital readmission. What is new?We developed an EHR-based deep learning model trained by 330 million data points from one of the largest cohorts of heart failure patients to date. Despite this model’s highly accurate prediction of long-term outcomes, such as mortality and disease progression, our findings suggest that EHR data alone offers limited predictive power for predicting the short-term outcomes of 30-day hospital readmission or death. What are the clinical implications?Our study supports the notion that hospital readmission, in contrast to other health outcomes, is uniquely driven by additional factors beyond traditional EHR variables. Once identified, incorporation of these determinants into future deep learning models could allow for accurate heart failure risk-stratification at hospital discharge to facilitate more efficient allocation of limited resources to the most vulnerable patients.Keywords
This publication has 40 references indexed in Scilit:
- Machine Learning–Based Models Incorporating Social Determinants of Health vs Traditional Models for Predicting In-Hospital Mortality in Patients With Heart FailureJAMA Cardiology, 2022
- Predicting machine's performance record using the stacked long short‐term memory (LSTM) neural networksJournal of Applied Clinical Medical Physics, 2022
- Tailored risk assessment of 90‐day acute heart failure readmission or all‐cause death to heart failure with preserved versus reduced ejection fractionClinical Cardiology, 2022
- Predicting 90 day acute heart failure readmission and death using machine learning‐supported decision analysisClinical Cardiology, 2020
- Machine learning and statistical methods for predicting mortality in heart failureHeart Failure Reviews, 2020
- Prediction of 30-Day All-Cause Readmissions in Patients Hospitalized for Heart FailureJAMA Cardiology, 2017
- A mixed-ensemble model for hospital readmissionArtificial Intelligence in Medicine, 2016
- Preventing 30-Day Hospital ReadmissionsJAMA Internal Medicine, 2014
- Diagnoses and Timing of 30-Day Readmissions After Hospitalization for Heart Failure, Acute Myocardial Infarction, or PneumoniaJAMA, 2013
- National and Regional Trends in Heart Failure Hospitalization and Mortality Rates for Medicare Beneficiaries, 1998-2008JAMA, 2011