Page 131 - Screening for Cervical Cancer: Systematic Evidence Review

Appendix C. Evidence Tables

Evidence Table 2. New Methods for Preparing or Evaluating Cervical Cytology (cont'd)

Patients & Methods Outcomes Measured Study Results & Limitations* Quality Considerations
Papnet
Women undergoing All Pap results: Prevalence Quality Score=9
routine screening aged abnormal defined as SCC=71/354,631=0.02% Ref. Std: 2
30 to 60 years CIN-I or greater Blind: 2
(194,358), or screening Estimated Se (CIN-I by Verification: 0
for any other reason Reference standard: Papnet+, SCC)=90.4% Consecutive: 2
(160,373). biopsy-confirmed Specificity cannot be Spectrum: 1
squamous cell calculated Publication: 1
carcinoma Industry: 1

No gold standard verification
of all test positives and test
negatives; specificity cannot
be calculated
ThinPrep 2000
462 subjects in first Concordance between Unable to get estimate of Quality Score=3.5
group, 1239 in following ThinPrep and specificity because no test Ref. Std: 0
group. conventional smear; positives verified with Blind: 2
discrepancies and 5% histology. Verification: 0
All patients >18 years, of test-negatives Consecutive: 0
premenopausal, and no verified by single ThinPrep more likely to identify Spectrum: 1
history of abnormal Pap independent slides as abnormal (LSIL+) Publication: 1 Industry: .5
pathologist only for than conventional smear
second group

6747 women, recruited Concordance between Unable to get estimate of Quality Score=5.5
from 6 centers. 3 ThinPrep and specificity because no test Ref. Std: 0
hospital centers conventional smear; positives verified with Blind: 2
selected for high discrepancies and 5% histology. Verification: 1
proportion of high-risk of test-negatives Consecutive: 0
women. verified by single ThinPrep more likely to identify Spectrum: 1
independent slides as abnormal (ASCUS or Publication: 1 Industry: .5
58.9% white, 21.7% pathologist LSIL) than conventional smear
black, 13.9 % Hispanic, in screening centers, but not in
2.2% other hospitals

C-37