Why the study?
A heart failure readmission prediction model using the Nationwide Readmissions Database was needed to guide clinical management and reduce heart failure readmissions.
Population
Approximately 1,420,564 patients admitted for HF in the Nationwide Readmissions Database
Comparison
Derivation cohort (2010-2012) vs validation cohort (2013-2014)
Design
Retrospective cohort analysis
Follow-up
180 days
Key result
A predictive model for heart failure readmission using the Nationwide Readmissions Database yielded an area under the ROC curve of 0.58 in the derivation cohort and 0.59 in the validation cohort.
Authors
Loading...
Limited discrimination cautions against clinical use for readmission forecasting; leaves open development of improved models with richer predictors.
Cohort (n=1,420,564)
Yes
A predictive model for heart failure readmission using a nationwide database demonstrated poor discriminative ability with an AUC of 0.58-0.59.
Zheng et al. (2022) conducted a cohort in Heart failure (n=1,420,564). Heart failure readmission prediction model was evaluated on Heart failure readmission. A predictive model for heart failure readmission using the Nationwide Readmissions Database yielded an area under the ROC curve of 0.58 in the derivation cohort and 0.59 in the validation cohort.