Synapse
⌘+K
Synapse
PulseExploreJournal ClubResearchersJournals
Instagram
HomeJournal ClubExplore
July 12, 2026JAMIA OpenOpen Access

Build fair machine learning models to predict adverse outcomes for heart failure patients with preserved ejection fraction and with reduced ejection fraction

View Full Paper
Ask AI
Bookmark
Share

Why the study?

To develop and validate subtype-specific, fairness-aware ML models integrating clinical and SDoH information to predict 6-month adverse outcomes in HFpEF or HFrEF and assess demographic subgroup error disparities.

Does integrating social determinants of health into machine learning models improve the prediction of 6-month readmission or mortality in patients with HFpEF or HFrEF?

Population

Adult HF hospitalizations with HFpEF or HFrEF from electronic health record data (2016-2022)

Comparison

Models with clinical and SDoH features vs clinical characteristics alone

Design

Retrospective EHR-based cohort study for ML model development and validation

Follow-up

6 months

Key result

Integrating social determinants of health into machine learning models modestly improved the C statistic for 6-month readmission or mortality in HFpEF (0.603 vs 0.586) and HFrEF (0.641 vs 0.637).

Authors

YYY. YehYLYao An LeeYHYu Huang

Discussion

Loading...

Member takes

Overview

May enhance equitable HF risk stratification with SDOH integration; leaves open prospective validation before clinical adoption.

Key Points

  • This research aims to create and validate machine learning models that include social determinants of health to predict 6-month readmission or mortality in heart failure patients.
  • Used electronic health record data from 2016-2022 to identify heart failure hospitalizations.
  • Trained logistic regression and XGBoost models with clinical and social determinants of health features.
  • Evaluated model performance and fairness across demographic groups using various statistical methods.
  • Adding social determinants of health increased the C statistic for logistic regression: HFpEF (0.603 vs 0.586) and HFrEF (0.641 vs 0.637).
  • FNR ratios indicated racial disparities; HFpEF FNRBlack/FNRWhite was 0.7834, improved to 0.8728 after debiasing techniques.
  • SHAP identified key predictive factors such as sodium levels and financial constraints in both heart failure subtypes.

Study Design

Type

Observational

Multicenter

No

Structured PICO

Does integrating social determinants of health into machine learning models improve the prediction of 6-month readmission or mortality in patients with HFpEF or HFrEF?

P
Population
Adult patients hospitalized with heart failure (HFpEF or HFrEF) followed for 6 months.
E
Exposure
Machine learning models (logistic regression and XGBoost) integrating clinical and social determinants of health (SDoH) information
C
Comparator
Machine learning models using clinical characteristics alone
O
Outcome
Composite outcome of readmission or mortality at 6 monthscomposite

Integrating social determinants of health into machine learning models provides modest discrimination gains and enables subgroup fairness assessment for predicting 6-month outcomes in heart failure patients.

Cite This Study

Yeh et al. (2026) conducted an observational in Heart failure (HFpEF and HFrEF). Integration of social determinants of health (SDoH) into machine learning models vs. Models without SDoH was evaluated on Composite outcome of 6-month readmission or mortality. Integrating social determinants of health into machine learning models modestly improved the C statistic for 6-month readmission or mortality in HFpEF (0.603 vs 0.586) and HFrEF (0.641 vs 0.637).

synapsesocial.com/papers/6a534ff4985393734a594cc6https://doi.org/10.1093/jamiaopen/ooag136
View Full Paper
Ask AI
Bookmark
Share

Also Consider

Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context:

  1. 1Fairness-Aware Machine Learning for Heart Failure Prediction: Performance, Bias, and Clinical Deployment Insights2025
  2. 2Comparison of Machine Learning Algorithms for Predicting Hospital Readmissions and Worsening Heart Failure Events in Patients With Heart Failure With Reduced Ejection Fraction: Modeling Study2023 · 30 citations
  3. 3Comparison of Machine Learning Algorithms for Predicting Hospital Readmissions and Worsening Heart Failure Events in Patients With Heart Failure With Reduced Ejection Fraction: Modeling Study (Preprint)2022
  4. 4Machine learning models in heart failure with mildly reduced ejection fraction patients2022 · 11 citations
  5. 5Beyond Composite Indices: Comprehensive Social Determinants Improve Heart Failure Readmission Prediction2026 · 1 citations