Skip to main content
. Author manuscript; available in PMC: 2017 Apr 20.
Published in final edited form as: JAMA Otolaryngol Head Neck Surg. 2017 Apr 1;143(4):382–388. doi: 10.1001/jamaoto.2016.3795

Table 3.

Multivariable Analysisa of Patient, Tumor, and Treatment Variables and Their Association With Ultimate Locoregional Control

Covariate Odds Ratio (95% CI) P Value
Lymphovascular space invasion
 Positive 0.22 (0.08–0.59) <.01
 Negative 1 [Reference]
Margins
 Positive 0.69 (0.06–7.72)   .76
 Close, <5 mm 0.24 (0.09–0.62) <.01
 Negative 1 [Reference]
Use of elective neck dissection
 Yes 1 [Reference]
 No 2.25 (0.89–5.68)   .09
a

Backward selection with an α level of removal of .20 was used. The following variables were removed from the model: age, histological grade, depth of invasion, perineural invasion, race, sex, adjuvant therapy, and T-stage. Having ultimate locoregional control was the reference group.