151751-Najiba-Chargi

116 CHAPTER 7 STATISTICAL ANALYSIS Data analyses was performed using IBMSPSS statistics 25. Descriptive statistics for continuous variables with a normal distribution were presented as mean with standard deviation (SD). Variables with a skewed distribution were presented as median with interquartile range (IQR). Categorical variables were presented as frequencies and percentages. Imputation analysis was performed in case of missing clinical variables. Logistic regression was used for univariate and multivariate analysis of surgical complications, only patients with known SMM status were included for analysis. Cox proportional hazard regression model was used for univariate and multivariate analysis of survival, only patients with known SMM status were included for analy- sis. Covariates used in the multivariate analysis were selected based on clinical significance or selected based on statistical significance (p < 0.05) in univariate analysis. Correlation analysis was performed by use of Pearson’s correlation analysis for variables with a normal distribution and Spearman’s correlation analysis was used for non-normally distributed variables. In case of high multicollinearity of variables in the multivariate analysis, highly correlated predictors were not included to prevent biased estimation. 33 Statistical significance was evaluated at the 0.05 level using two-sided tests. Survival was visualized using Kaplan Meier survival curves. RESULTS PATIENT CHARACTERISTICS Descriptive data are described in table 1. In total, 616 patients were included. Median age at diagnosis was 60.8 years (IQR 51.6-69.5). Of these patients, 554 patients (89.9%) were onco - logical cases of which 509 patients (91.9%) were diagnosed with head and neck squamous cell carcinoma (HNSCC). Majority of patients were male (60.7%). Most used flap was the radial forearm free flap (RFFF) (n=276, 44.8%). Figure 1. shows the flaps used for reconstruction.