Segmental strain in chronic ischemic cardiomyopathy: clarifying the role of multiple-comparison adjustment and the clinical interpretation of limits of agreement

Khidr and colleagues recently reported an intermodality comparison of left ventricular longitudinal strain by feature-tracking cardiac magnetic resonance (FT-CMR) and 2D speckle-tracking echocardiography (2D-STE) in 55 stable patients with chronic ischemic heart disease and reduced ejection fraction. Their analysis showed a strong global longitudinal strain (GLS) correlation (r = 0.793, P < 0.001) and explicitly cautioned against treating the two modalities as interchangeable for individual or longitudinal assessment. We support these conclusions and write to extend two methodological points. First, the 17 segment-level paired comparisons in Table 4 are reported without explicit multiplicity adjustment. On recomputation against a Bonferroni threshold of α/17≈0.00294, 11 of 17 segmental differences remain statistically significant, indicating that the original segmental signal is more statistically robust than an unadjusted reader might assume; we therefore recommend explicit reporting of adjusted P-values together with a unified mixed-effects formulation that accommodates within-patient segment-to-segment correlation. Second, the GLS limits of agreement span 8.3 percentage points (− 3.2% to + 5.1%), a width sufficient to misclassify individual patients across established GLS thresholds. A worked example is provided, and the recommendation to confine longitudinal follow-up to a single platform is argued to deserve more prominent placement in the Abstract and Conclusions.