A secondary diagnostic accuracy analysis of 34 patients found that a global longitudinal strain (GLS) cut-off of -17.0% identified CMR-defined reduced left ventricular ejection fraction (LVEF <50%) with an AUC of 0.892, sensitivity of 75%, and specificity of 86.4% — outperforming speckle-tracking-derived LVEF (AUC 0.771, sensitivity 58.3%, specificity 95.5%), though the difference did not reach statistical significance (DeLong p=0.192). Twelve of 34 patients had reduced LVEF by cardiac MRI, the recognized gold standard.
The clinical appeal here is real: cardiac MRI remains expensive, time-consuming, and inaccessible in many settings, making reliable echocardiographic surrogates valuable for triaging patients with suspected systolic dysfunction. GLS has attracted growing interest precisely because it detects subtle myocardial dysfunction before LVEF visibly declines — a phenomenon well-documented in chemotherapy-related cardiotoxicity and early cardiomyopathy screening. A validated GLS threshold aligned to CMR-defined reduced LVEF would meaningfully streamline referral decisions.
However, the limitations here are substantial. The cohort of 34 patients — only 12 with reduced LVEF — is far too small to anchor a clinical cut-off with confidence; confidence intervals spanning 0.538–0.996 signal instability. This is a single-center, cross-sectional, secondary analysis, meaning selection bias cannot be excluded. Critically, this is a preprint posted on medRxiv and has not yet undergone peer review — conclusions may shift materially. The authors themselves correctly characterize findings as exploratory and hypothesis-generating, warranting multicenter validation before any clinical adoption.