Intra- and inter-rater reproducibility of the echocardiographic data of the Nigerian UPRIGHT-HTM trial participants

Abstract

Background: Echocardiography is the guideline- recommended tool for risk stratification in asymptomatic high-risk patients, but its reproducibility in sub-Saharan Africa remains largely undocumented. This study assessed the intra- and inter-rater repeatability of the risk-carrying echocardiographic traits among 42 patients enrolled in the UPRIGHT-HTM trial at three Nigerian sites. Methods: The echocardiographic images (168 for intra- and 84 for inter-rater agreement) were acquired according to current recommendations and blindly analysed by randomising the order of the digitised images. The study focused on left ventricular mass (LVM) index to body surface area of height 2.7 and ejection fraction (EF) and left ventricular diastolic dysfunction (LVDD). Repeatability was assessed by the Bland and Altman approach, the coefficient of variation (CV), the intra-class correlation coefficient (ICC) and Cohen's k statistic. The current findings were compared with a systematic literature review of 27 publications. Results: LVM showed a slight but significant intra-rater bias, but no inter-rater bias. CVs for LVM were approximately 30%, ICCs ranged from 0.74 to 0.84, and k statistics for LV hypertrophy (LVH) varied from 0.49 to 0.64. For EF, intra- and inter-rater CVs ranged from 13.0% to 14.5%, ICCs from 0.60 to 0.69, and statistics were 0.71 and 0.53, respectively. LVDD reproducibility was moderate, with κ values around 0.50. These findings were consistent with most of the reviewed literature. Conclusion: The risk-carrying echocardiographic traits were obtained with moderate repeatability in three Nigerian tertiary referral centres. As data quality augments with expertise, these findings call for a Nigerian training programme in echocardiography with certification.