Interpretation of cardiac chamber size on canine thoracic radiographs is limited and may result in the false identification of right-sided cardiomegaly in the presence of severe left-sided cardiomegaly.

Qualitative assessment of individual cardiac chamber enlargement on thoracic radiographs was assessed using echocardiography as the gold standard in dogs. Using this method, the presence of severe left-sided cardiomegaly was hypothesized to contribute to the false identification of right-sided cardiomegaly on radiographs. 175 dogs with three-view thoracic radiographs were retrospectively included in this diagnostic accuracy study if echocardiography was done within 24 h, and no rescue therapy was provided in the interim. All radiographic studies were reviewed by two groups of five board-certified veterinary radiologists with greater and less than 10 years of experience for grading of cardiac chamber enlargement as normal or mildly, moderately, or severely enlarged. The agreement, sensitivity, and specificity of the radiologists' interpretation of cardiac chamber size on thoracic radiographs to measured echocardiographic grades were evaluated. A total of 147 cases had complete echocardiographic data available for analysis. Intragroup agreement was moderate for the evaluation of left atrial enlargement and slight to fair for all other cardiac chambers. Between the mode of the radiologists' responses in the two groups and the echocardiographic grades, there was slight agreement for all cardiac chambers with higher severity grades reported using echocardiography. When moderate to severe left-sided cardiomegaly was identified on echocardiography, the sensitivity, specificity, and accuracy of radiographs were low, identifying dogs with radiographic evidence of right-sided cardiomegaly in the absence of corresponding right-sided cardiomegaly on echocardiography. Therefore, thoracic radiographs should be used with caution for the evaluation of cardiac chamber enlargement, particularly in the presence of severe left-sided cardiomegaly.