We recently published in JFP a meta-analysis of liquid-based cervical cytology (LBCC) tests.1 Because these technologies are not identical, we tested the hypothesis that there is a difference between ThinPrep and other LBCC tests. We therefore repeated our analysis using only the 3 studies that compared both ThinPrep (TP) and conventional Pap smears with a reference standard.
In this repeat analysis the pooled sensitivity (0.81) and specificity (0.81) for ThinPrep were lower than in the combined analysis for all LBCC tests. However, the sensitivity and specificity for conventional Pap smear were also lower; the net difference in sensitivity between TP and conventional Pap smear is therefore only 1% greater than we initially reported. Unfortunately, the lack of an adequate reference standard in most of the existing studies does not allow for a true comparison of the 2 tests, and the studies with reference standards reveal relatively small differences that fail to reach statistical significance.
With new guidelines for treatment of ASC-US recommending triage with HPV DNA tests, the convenience of using the ThinPrep sample already obtained for reflex testing is an advantage. Use of the ThinPrep for repeat testing of women with ASC-US Pap tests to facilitate HPV testing is an option as well. More studies are needed to compare these new strategies with traditional approaches to the management of these patients.
Sandra M. Sulik, MD, MS
Lorne A. Becker, MD
William D. Grant, EdD
Upstate Medical University
State University of New York
Syracuse
E-mail: smsulik@aol.com