Spss 13

All statistical analyses were performed using SPSS 13 (IBM, Armonk, NY). The differences in categorical variables such as age, sex, tumor location, size, gross pattern, depth of invasion, differentiation, micropapillary adenocarcinoma, LVI, LNM, and perineural invasion were calculated between the groups using the χ², Fisher exact test, or Kruskal-Wallis H test. In multivariate analysis, the odds ratio (OR) with the confidence interval (CI) at the 95% level was calculated, and P value <0.05 was defined as the critical value of the evaluation factor included in the final risk model. The assignment of points to risk factors was based on a linear transformation of the corresponding β regression coefficient. The coefficient of each variable was divided by the lowest β value and rounded to the nearest integer (18 (link)). The scoring points of the patients were added and divided into 3 LNM risk categories: low risk (<3%), intermediate risk (3%–19.6%), and high risk (>19.6%) (18 (link)). The classification models were compared based on the area under the receiver operating characteristic curve (AUC-ROC). The value of this metric ranges from 0 to 1, and a higher score is preferred.

All the statistical analyses were done using IBM SPSS 13. The distribution of categorical variables such as gender, diagnosis, indication of intubation, etc. was expressed as frequency and percentages. The continuous data such as age, duration of intubation, FOIS score, salivary flow, etc. were expressed as mean with standard deviation or median with interquartile range (IQR). Comparison of baseline characteristics between the groups was done using the Chi-square test, Fisher’s exact test, Mann Whitney U test, and Independent Student t-test. The longitudinal change of salivary flow and oral intake status were compared between groups using Repeated Measures Analysis of Variance (RMANOVA) and Mann Whitney-U test respectively.