These tools will no longer be maintained as of December 31, 2024. Archived website can be found here. PubMed4Hh GitHub repository can be found here. Contact NLM Customer Service if you have questions.


Pubmed for Handhelds

PUBMED FOR HANDHELDS

Search MEDLINE/PubMed

  • Title: Application of logistic regression in combination with multiple diagnostic tests for auxiliary diagnosis of nasopharyngeal carcinoma.
    Author: Jiang SQ, Liu Q.
    Journal: Ai Zheng; 2009 Feb; 28(2):177-80. PubMed ID: 19550133.
    Abstract:
    BACKGROUND AND OBJECTIVE: Although there are many markers for the clinical diagnosis of nasopharyngeal carcinoma (NPC), the efficacy of most of the markers for the early diagnosis is poor. This study was to evaluate the diagnostic value of VCA/IgA, EA/IgA, EBV DNA, EBNA1/IgA, EBNA1/IgG and ZTA/IgG for NPC, as well as to screen out an optimized combination using the logistic regression to increase diagnostic accuracy of NPC. METHODS: Eight-one newly pathologically diagnosed NPC patients prior to treatment and 89 health cases from routine physical checkups were entered into the study. Epstein-Barr virus (EBV) DNA was detected by quantitative real-time PCR; VCA/IgA and EA/IgA were assessed by immunofluorescence assays (IFA). The receiver operating characteristic (ROC) curve and the area under the curve were used to evaluate the diagnostic value of a single test or combined tests for NPC, thus to decide the cut-off value. The logistic regression model was used to combine the results from multiple tests to increase diagnostic accuracy. RESULTS: Comparing to the routine parallel sequential test, the logistic regression in combination with multiple diagnostic tests achieved higher diagnostic specificity and sensitivity for NPC. Two optimal combinations were EBV DNA + EBNA1/IgA and VCA/IgA + EBNA1/IgA, whose sensitivity and specificity reached 0.96 and 0.82, 1.00 and 0.84, respectively. When the logistic model was used and the cut-off value was determined by ROC, the sensitivity and specificity of the two combination groups became 1.00 and 0.87, 0.98 and 0.88, respectively. CONCLUSION: Adopting the logistic regression in combination with multiple diagnostic tests and using the probability prediction to decide the cut-off value may help increase the diagnostic sensitivity and specificity for NPC.
    [Abstract] [Full Text] [Related] [New Search]