Clinical–core lab discrepancies in echocardiography and their impact on implantable cardioverter-defibrillator eligibility in hypertrophic cardiomyopathy
B Chen, I Henarejos Castillo, C Henri, J Boulet, P Garceau, P L. L'Allier, G Giraldeau, D Lee, G Oudit, H Khan, T M Roston, C Steinberg, R Tadros, J C Tardif, M Tremblay-GravelAbstract
Background
Accurate transthoracic echocardiographic (TTE) assessment is central to the diagnosis and risk stratification of hypertrophic cardiomyopathy (HCM), yet interobserver differences exist due to complex geometry and hemodynamics. While core lab assessments are standard in research, clinical decision-making relies on local interpretations. The level of agreement between clinical and core lab measurements has not been systematically evaluated, nor has the impact of any discrepancies on implantable cardioverter-defibrillator (ICD) implantation decisions.
Purpose
To evaluate the agreement between clinical and core lab TTE measurements in HCM and its impact on primary prevention ICD decision-making to clarify the role of core lab reading in research and its implications for clinical practice.
Methods
The HiRO-HCM registry provided TTE clinical reports and DICOM images from patients with HCM or unaffected variant carriers across 14 Canadian hospitals. Images were transferred to a core lab for centralized analysis. Agreement between clinical and core lab measurements of maximal left ventricular wall thickness (MLVWT), left ventricular ejection fraction (LVEF), maximal left ventricular outflow tract (LVOT) gradient and left atrial diameter (LAD) was evaluated using scatter plots, Bland–Altman plots with 95% limits of agreement, and intraclass correlation coefficients (ICC). ICD insertion recommendations were assessed using the European Society of Cardiology (ESC) 5-year risk score for sudden cardiac death (SCD) in HCM and the American Heart Association (AHA) guidelines, for both clinical and core lab measurements; this analysis was restricted to one TTE per patient and studies with complete clinical data.
Results
Agreement analyses included 2,899 TTE studies from 1,518 adult patients (Figure 1). Agreement was moderate for MLVWT (ICC = 0.69) and LVEF (ICC = 0.62), and strong for maximal LVOT gradient (ICC = 0.91) and LAD (ICC = 0.85). Core lab intra- and interobserver ICC values for LVWT, LVEF and LAD were between 0.92 and 0.99. On Bland–Altman analysis, mean bias (clinical report – core lab) was negative (clinical report underestimation) for MLVWT and LVEF, and positive (clinical report overestimation) for LVOT gradient and LAD. ICD impact analyses included 594 patients (Table 1). Comparison for the ESC score showed 1 out of 594 (0.2%) reclassified across two risk categories (high to low or low to high) between clinical and core lab assessments. Comparison for the AHA guidelines showed 36 out of 594 (6.1%) disagreeing for Class IIa ICD implantation.
Conclusion
Clinical and core laboratory measurements showed moderate to strong agreement; however, variability near decision thresholds may influence ICD eligibility, particularly under AHA criteria. Confirmatory imaging may therefore be warranted when measurements approach guideline cutoffs, while centralized core laboratory analysis enhances research reliability.Clinical–Core Lab Agreement (MLVWT, LVEF)ESC/AHA ICD: Core Lab vs Clinical Reports