Abstract
The heart-to-single vertebra ratio (HSVR) has been proposed as a simple, quick, and reliable radiographic method to assess cardiac silhouette dimensions in dogs. The HSVR shows excellent agreement with the vertebral heart size (VHS), and it can also be accurately determined in dogs with vertebral abnormalities affecting the thoracic spine. This retrospective, single-center, method-comparison, observer-agreement study investigated the reliability of the HSVR in cats. Three observers retrospectively evaluated anonymized right-lateral thoracic radiographs obtained over a set period of time. Exclusion criteria included the presence of thoracic spine alterations and the inability to outline the cardiac silhouette. The HSVR was calculated by dividing the sum of the cardiac long and short axes by the length of each vertebral body from T4 to T8. Images of 101 cats of different breeds met the inclusion criteria. Lin's concordance correlation coefficient and the relative 95% confidence interval (CI) revealed that the HSVR(T6) showed the best agreement with the VHS (0.95; 95% CI: 0.91-0.97). Bland-Altman plots showed low bias (-0.023 ± 0.19; limits of agreement = -0.39 to 0.35) between the HSVR(T6) and the VHS, with low mean absolute error (0.14; 95% CI: 0.12-0.17) between the two methods. The intraclass correlation coefficients, evaluated on 20 cats, demonstrated excellent interobserver agreement (0.95-0.96; p < .001) and good to excellent intraobserver agreement (0.75-0.94; p < .001) for all HSVRs. The results of this study confirmed that the HSVR is a simple, quick, and reliable alternative to the VHS also in cats.