Key result
Unsupervised clustering of electronic health record data identified four distinct phenotypic heart failure subtypes in both males and females, which differed significantly by age, comorbidities, and structural abnormalities rather than ejection fraction.
Why the study?
High-dimensional electronic health records data can be used to phenotype complex diseases, prompting the application of unsupervised clustering to EHR-based traits in heart failure.
Observational (n=8,569)
No
Unsupervised machine learning applied to routine EHR data can accurately classify heart failure patients into distinct phenotypic subtypes, demonstrating that LVEF is not a primary driver of these clusters.
Clustering HF beyond LVEF via routine EHR data may refine subtypes; leaves open whether clusters predict outcomes or guide therapy.
Background High-dimensional electronic health records (EHR) data can be used to phenotype complex diseases. The aim of this study is to apply unsupervised clustering to EHR-based traits derived in a cohort of patients with heart failure (HF) from a large integrated health system. Methods Using the institutional EHR, we identified 8569 patients with HF and extracted 1263 EHR-based input features, including clinical, echocardiographic, and comorbidity data, prior to the time of HF diagnosis. Principal component analysis, Uniform Manifold Approximation and Projection, and spectral clustering were applied to the input features after sex stratification of the cohort. The optimal number of clusters for each sex-stratified group was selected by highest Silhouette score and by within-cluster and between-cluster sums of squares. Determinants of cluster assignment were evaluated. Results We identified four clusters in each of the female-only (44%) and male-only (56%) cohorts. Sex-specific cohorts differed significantly by age of HF diagnosis, left ventricular chamber size, markers of renal and hepatic function, and comorbidity burden (all p<0.001). Left ventricular ejection fraction was not a strong driver of cluster assignment. Conclusion Readily available EHR data collected in the course of routine care can be leveraged to accurately classify patients into major phenotypic HF subtypes using data driven approaches.
No takes yet. Share an insight, caveat, or question.
Reza et al. (2022) conducted an observational in Heart failure (n=8,569). Unsupervised clustering of EHR-derived phenotypes was evaluated on Identification of phenotypic HF subtypes (clusters) and intercluster differences (p=<0.001). Unsupervised clustering of electronic health record data identified four distinct phenotypic heart failure subtypes in both males and females, which differed significantly by age, comorbidities, and structural abnormalities rather than ejection fraction.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: