Diagnosing diastolic dysfunction using explainable advanced electrocardiography

Abstract

AbstractBACKGROUND AND AIMSDiastolic dysfunction of the left ventricle is a precursor to heart failure with preserved ejection fraction (HFpEF). Detection by electrocardiography (ECG) when asymptomatic (stage B heart failure) would be valuable. We hypothesised that an explainable advanced ECG (A-ECG) score derived from standard 12-lead ECG could accurately diagnose diastolic dysfunction.METHODSIncluded patients had undergone resting 12-lead ECG and echocardiography demonstrating normal systolic function, with at most mild valve disease, and either the presence (n=150) or absence (n=264) of grade II or III diastolic dysfunction. Stepwise multivariable logistic regression was used to generate an A-ECG score that was crossvalidated using bootstrapping.RESULTSA 6-measure A-ECG score was able to identify diastolic dysfunction with an area under the receiver operating characteristics curve [95% confidence interval] of 0.91 [0.88-0.94], sensitivity 83 [76-91]%, specificity 87 [77-92]%, positive predictive value 78 [69-85]%, negative predictive value 90 [86-94]%, positive likelihood ratio 6.38 [3.30-11.38] and inverse negative likelihood ratio 5.12 [3.21-10.22].CONCLUSIONSStandard 12-lead ECG can be used to accurately identify diastolic dysfunction by echocardiography via A-ECG. This may have clinical utility for early identification of patients who may benefit from further cardiac assessment and risk factor management to prevent progression to symptomatic heart failure.