Statistical Analysis

All statistical analyses were performed using commercially available software (R 2.8.1). Categorical data are expressed as absolute numbers and frequencies as percentages. Continuous data are reported as median (interquartile range).

Shapiro–Wilk test was performed to evaluate the normality of the data. Categorical indices were compared among the PH probability groups using Fisher's exact test. Continuous indices were compared among the PH probability groups using one-way analysis of variance with subsequent pairwise comparisons using Tukey's multiple comparison test for normally distributed data or the Kruskal–Wallis test with subsequent pairwise comparisons using the Steel–Dwass test for non-normally distributed data. All echocardiographic indices for right heart morphology and function were compared for the presence or absence of R-CHF using a Student's t-test for normally distributed data or the Mann–Whitney U test for non-normally distributed data. Additionally, univariate and multivariate logistic regression analyses were performed to evaluate the association between the presence of R-CHF and echocardiographic indices for the right heart, including PVRecho. After adjusting for multicollinearity, variables with P < 0.10 in the univariate analysis, were entered into the multivariate analysis. Results of logistic regression analyses were recorded as adjusted odds ratios and their respective 95% confidence intervals (CI). Receiver operating characteristic curves were created to calculate the area under the curve (AUC), sensitivity, and specificity, and to determine the optimal cutoff values required for evaluating the presence of R-CHF. The AUC was considered to have high accuracy if it was >0.9, moderate accuracy if it was 0.7–0.9, and low accuracy if it was 0.5–0.7 (30). The optimal cutoff value was defined as the value that minimized the distance between the curve and the upper left corner in the receiver operating characteristic curve.

Intra-observer variability measurements were performed by a single observer who performed all echocardiographic measurements (YY). The echocardiographic indices assessed in this study were obtained from nine dogs (three dogs in each PH probability group). All measurements were performed on two different days with >7-day intervals using the same cardiogram and cardiac cycles. A second blinded observer (HKa) measured the same indices to determine interobserver variability using the same echocardiogram and heart cycles. Variability of intra- and inter-observer measurements was quantified by the coefficient of variation (CV), which was calculated using the following formula:

CV (%)=(standard deviation)(mean value)×100

Intra- and inter-class correlation coefficients (ICCs) were also used to evaluate measurement variability. Low measurement variability was defined as CV < 10.0, and ICC > 0.70.

Statistical significance was set at P < 0.050 for all the analyses.