Skip to main content

Table 4 Cox proportional hazard models for predicting overall survival (N = 60)

From: Clinical outcomes of bulky parotid gland cancers: need for self-examination and screening program for early diagnosis of parotid tumors

Variables Category Univariable analysis Multivariable analysis #1c Multivariable analysis #2c
HR 95%CI P HR 95%CI P HR 95%CI P
Age Continuous 1.040 1.009 1.073 0.012 1.013 0.977 1.050 0.492 1.035 1.000 1.071 0.051
Sex M vs. F (Ref) 3.157 0.932 10.690 0.065 1.481 0.403 5.448 0.554 1.309 0.350 4.901 0.689
Facial nerve palsy at diagnosis Yes vs. No (Ref) 1.541 0.645 3.679 0.330         
Tumor sizea Continuous 1.250 1.065 1.467 0.006 1.215 1.029 1.435 0.022 1.185 1.005 1.398 0.044
Node metastasisb Yes vs. No (Ref) 4.586 1.783 11.800 0.002      3.998 1.428 11.193 0.008
Tumor grade High grade vs Others (Ref) 3.353 1.382 8.135 0.007 5.385 1.657 17.499 0.005     
Perineural invasion Yes vs. No (Ref) 1.373 0.533 3.534 0.511         
Surgery Yes vs. No (Ref) 0.408 0.171 0.975 0.044 0.255 0.080 0.817 0.021 0.932 0.332 2.613 0.893
Distant metastasis Yes vs. No (Ref) 2.098 0.704 6.256 0.184         
  1. aTumor size: Determined by preoperative radiology
  2. bNode metastasis: Clinical nodal status
  3. We used clinical variables (tumor size and cN) in this analysis, because 11 patients did not undergo surgery
  4. cDue to a possibility of potential correlation between lymph node metastasis and tumor grade (high grade) (P = 0.05), we separately constructed two multivariable models
  5. HR Hazard ratio, 95%CI 95% confidence interval