Statistical Analysis
Baseline characteristics were presented as mean ±SD for continuous variables and compared using the Student t -test, or percentages for categorical variables differences compared using the chi-square test. A p-value < 0.05 was defined as statistically significant. Univariate and multivariate analysis based on the logistic regression model were performed to determine the TTE parameters to estimate the elevated LV filling pressure. Only variables with p value< 0.05 in univariate analysis were entered into multivariate analysis. Correlation between LV GLS and diastolic parameters were analyzed using the Pearson correlation method. Correlation of invasive LV filling pressure with LV GLS and diastolic parameters were also analyzed using the Pearson correlation method. Sensitivity, specificity, positive predictive value, and negative predictive value of diastolic parameters and LV GLS were analyzed using the Receiver operating characteristic (ROC) analysis based on the Logistic regression method. LV GLS cut-off value was determined by ROC analysis. All data were analyzed using JMP version 14.0 (SAS Institute Inc., Cary, North Carolina)
Inter-observer and intra-observer variability . Images from 10 patients were randomly selected, and a second independent blinded observer measured their images to assesses the inter-observer variability. The first observer who measured all patients’ views remeasured the same randomly selected 10 patients’ views at least 6 weeks apart from the first measurement. Inter-observer and intra-observer variability were assessed using the Intra Class Correlation Coefficient (ICC) method.