2.4. Statistical Analysis

Normally distributed continuous variables were expressed as mean ± standard deviation (SD) and compared between groups using analysis of variance. Skewed data were expressed as median and interquartile range (IQR) and were compared using the Kruskal-Wallis test. Categorical variables were expressed as number (percentage) and compared using the Chi-square test.

The relationship between PEf size and in-hospital mortality was identified by binary logistic regression analysis. We carried out multivariable analyses for the primary outcome. We entered all variables into the logistic analysis based on the stepwise regression method and clinical doubt. We selected the variables which were significantly associated with the outcome (p<0.05) and not correlated (Spearman correlation >0.5) in a logistic regression analysis. The results were expressed as odds ratio (OR) and 95% confidence interval (CI). The p for trend was calculated. The curve that conformed to the general trend was plotted through local weighted regression (Lowess). A two-tailed p<0.05 was considered statistically significant. All data analysis was performed by Stata V.15.1 (Statistical Analysis System, North Carolina, the United States).