Predicting the probability of malignancy of the neck mass with logistic regression model: a statistical analysis of excisional biopsy of neck masses

Zhonghua Yi Xue Za Zhi (Taipei). 1992 May;49(5):328-34.

Abstract

One hundred and fifty-six patients with neck lesions were selected into this retrospective study between January, 1989 and December, 1990. All the patients visited OPD with the chief complaint of neck mass without other apparent symptoms and signs after initial work-up. They can be divided into 5 types according to the pathological reports: type a) metastatic lesion, type b) malignant lymphoma, type c) TB lymphadenitis, type d) miscellaneous benign lesion, and type e) inadequate specimen. They represent 24.4%, 12.8%, 9.6%, 52.6%, and 0.6% of the patients respectively. Type a) and type b) were classified as group of malignancy and the other three types were group of benignancy. Chi-square test and t-test were then used to evaluate the significant level of difference in each semiological factor between both groups. Several parameters were found to reach significant level, including tumor fixation, age of the patients, tenderness, location and size of the tumor, and history of cancer. Stepwise logistic regression was utilized to obtain a regression equation to predict the probability of malignancy in OPD patients with neck masses. The accuracy rate of prediction is 83.3%, if the cutpoint value is 0.5. A clinician can therefore avoid untimely excisional biopsy and delay in treatment planning.

MeSH terms

  • Adolescent
  • Adult
  • Aged
  • Aged, 80 and over
  • Biopsy
  • Child
  • Child, Preschool
  • Female
  • Head and Neck Neoplasms / diagnosis*
  • Humans
  • Male
  • Middle Aged
  • Models, Statistical*
  • Neck / pathology*
  • Probability
  • Regression Analysis