SciELO journals
Browse
1/1
3 files

Analysis of the application values of different combination schemes of liquid-based cytology and high-risk human papilloma virus test in the screening of high-grade cervical lesions

dataset
posted on 2018-11-28, 02:43 authored by Jian Wang

The aim of this study was to explore the value of different combination schemes of liquid-based cytology (LBC) and high-risk human papilloma virus (HPV) test in the screening of high-grade (≥CIN 2) cervical lesions. From 5727 women who had undergone examinations with LBC and high-risk HPV test, 1884 patients with positive results of either or both LBC and HPV test were included in this study and underwent cervical biopsy. Based on the pathological examination results, comparisons of the assessment indicators of all diagnostic tests were made, and the application values of LBC and high-risk HPV test and different combination schemes of the two in the screening of high-grade (≥CIN II) cervical lesions were estimated. Compared with the single test method, the sensitivity and negative predictive value of the combination scheme of LBC+HPV (with one positive result) were increased significantly (98.7% and 99.7%), but the specificity (60.8%) and accuracy rate (65.4%) dropped significantly (P<0.05). The sensitivity of LBC+HPV (with two positive results) was the lowest (80.7%), but the specificity and accuracy rate were the highest (83.5% and 83.1%, P<0.05). Z test showed that differences in the screening efficiency of four schemes were not statistically significant (P>0.05). Both LBC and HPV test were effective methods in the screening of high-grade cervical lesions; combination of the two tests did not improve the screening efficiency, but the scheme of LBC+HPV (with two positive results) significantly increased the sensitivity and negative predictive value, which was of better cost-benefit value.

History

Usage metrics

    Brazilian Journal of Medical and Biological Research

    Licence

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC