Spss version 22.0 software package

The IBM SPSS Version 22.0 software package was used for statistical computation (IBM Corp, Armonk, NY, USA). Survival estimates were calculated by the Kaplan-Meier product-limit method and compared with the log-rank test. Multivariate analysis was performed using a Cox regression model [22 ]. Also a Fine and Gray competing risk survival regression analysis was performed in R (version 3.5.2) using the library package “survival” and “cmprsk” to consider deaths from other causes as competing events. A p < 0.05 value was considered statistically significant. The impact of RT technique (3D-CRT vs IMRT/VMAT), ADT (yes or not), and PNI (yes or not) on the incidence of second primary tumors was estimated. Second tumors incidence was evaluated not only as “any second tumor” detected during the follow-up but also considering other 2 groups: i) second tumors in the pelvis and ii) second tumors in the abdomen or pelvis. In cases of doubtful interpretation of the information contained in the database for the purposes of this stratification, the diagnostic images of the second tumor were analysed.

All quantitative data are described as the means ± standard deviation (SD), and qualitative data are depicted as numbers (n) and percentages (%). Quantitative data was compared by Student’s t-test and qualitative data was compared by the χ² test or Fisher exact test. Factors related to study endpoints in univariate analysis and important clinical parameters were put into backward logistic regression analysis or Cox regression analysis to seek the independent predictors. Relation analysis and collinearity diagnosis were also performed, and only one of the variables with a significant correlation was included in the multivariate regression analysis. Similar to our previous studies, the Kaplan–Meier method was utilized to evaluate cumulative survival and the log-rank test was used to evaluate the differences between the two groups in the cumulative survival[15 (link)]. To further evaluate the effect of time to peak TB concentration on predicting patient in-hospital mortality, the area under curve of the receiver operating characteristic (AUC-ROC) was computed and the Youden index was employed to assess the optimal cutoff values of time to peak TB concentration. Bilateral P < 0.05 was regarded as statistically significant for all analyses. Statistical analysis was performed by using the IBM SPSS version 22.0 software package (SPSS Chicago, IL).