Page 60 - Screening for Cervical Cancer: Systematic Evidence Review

Chapter III. Results

having routine cervical cancer screening at 40 general practitioner practices approximates

21
screening use in routine primary care practice in the United States.

HPV types consistently considered high risk and tested by the assays used in these studies

included types 16, 18, 31, 33, 35, 39, 45, 51, 52, 56, 58, 59, and 68. We have not considered

performance characteristics of tests that detect only low-risk HPV types. Basic performance

characteristics relating presence of high-risk HPV types to diagnoses are best summarized by

considering HSIL and LSIL as separate diagnostic thresholds. For the HSIL threshold, we mean

the ability of a positive HPV test to detect histology-proven HSIL or more severe lesions, such as

invasive cancer. For the LSIL threshold, we mean the ability of a positive HPV test to detect

histology-proven LSIL or more severe lesions, including HSIL.

HPV Screening for Detection of High-grade Cervical Changes

The sensitivity of a positive Hybrid Capture II test for high-risk HPV types for detecting

histology proven HSIL or greater ranged from 62% to 95%. 21,62-65 Specificity ranged from 41%

to 94%. 21,62-65 Details of test performance for detecting high-grade lesions, including predictive

values and likelihood ratios for positive and negative tests, for each HPV test modality examined

in these screening-based studies appear in Table 8.

Five of the 6 HPV screening studies used Hybrid Capture II. 21,61-64 To report aggregate

estimates of performance for detection of HSIL, we calculated summary performance measures

using the aggregate numbers of true positives, false negatives, true negatives, and false positives

in these studies, with a combined total enrollment of 6,793 participants. In aggregate, the

sensitivity of Hybrid Capture II for detecting HSIL was 84.4%; specificity was 78.7%; positive

predictive value, 23.4%; and negative predictive value, 98.5%.

60