Statistical Analysis
Descriptive statistics are presented using numbers and percentages, mean and standard deviation (SD) or median and interquartile range (IQR) when appropriate. Patients with missing or incomplete responses were excluded from all analyses.
The test-retest reliability of the VAS was analyzed with the intraclass correlation coefficient (ICC, two-way random effects model, absolute agreement, single measures). As a measure of validity, the VAS score was correlated to the FaCE total score.
Two multiple linear regression models were examined. First, the FaCE subscale scores were correlated with total FaCE score. The regression coefficient of each subscale represents the size of the subscale’s correlation with the total FaCE score, which should exactly reflect the number of questions contained in each subscale.
Second, a linear regression analysis between the FaCE subscales and a transformed VAS score was performed. The original VAS scores were inverted so that a higher score indicated better quality of life, as seen with FaCE scale scores. The sizes of the regression coefficients were examined and compared with those from the first regression analysis.
The presence of multicollinearity in the second analysis was checked for by looking at the correlation, tolerance, and variation inflation factor (VIF) between each variable. A tolerance larger than 0.2 and a VIF less than 10 was considered to indicate the absence of multicollinearity.
All analyses were performed in the Statistical Package for the Social Sciences (SPSS) version 26 (IBM Corporation, NY, USA).