Accuracy of visually estimated left heart size by echocardiography in dogs with myxomatous mitral valve disease.

Abstract

INTRODUCTION/OBJECTIVES: The aim of this study was to assess the correlation between visual estimation of left heart size and conventional echocardiographic measurements in dogs with myxomatous mitral valve disease (MMVD). ANIMALS, MATERIALS AND METHODS: Seventy dogs with various stages of MMVD were retrospectively enrolled. Five investigators (two cardiologists and three non-cardiologists) received brief training before visually evaluating the left atrium-to-aortic ratio (LA:AO) and the presence of left ventricular enlargement using right parasternal long-axis five-chamber and basal short-axis echocardiographic videos. Correlations between visually estimated and conventionally measured LA:AOs were assessed using linear regression and Spearman's rank correlation. Interobserver agreement was evaluated using the intraclass correlation coefficient. Agreement in identifying left ventricular enlargement was assessed using the Fleiss Kappa coefficient. RESULTS: A strong correlation was found between visual estimation and conventional measurements of LA:AO (r = 0.89; &#x3c1; = 0.90; P<0.001). Interobserver agreement for LA:AO visual estimation was good (intraclass correlation coefficient = 0.76; 95% confidence interval: 0.65-0.84). Agreement between visual and conventional evaluation of left ventricular size was moderate (Fleiss Kappa = 0.50). STUDY LIMITATIONS: Limitations include the use of high-quality images obtained by a cardiologist, the predominance of small-breed dogs, and the use of a non-standard imaging view for left ventricular internal dimension at end-diastole normalized to body weight calculation. CONCLUSIONS: Visual estimation demonstrated strong correlation with quantitative LA:AO measurements and moderate agreement for left ventricular size. It may be a useful tool in emergency or primary care settings when conventional echocardiography is not feasible.