Page 133 - Screening for Cervical Cancer: Systematic Evidence Review

Appendix C. Evidence Tables

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

Study Results &
Patients & Methods Outcomes Measured Limitations* Quality Considerations
ThinPrep 2000
35,560 pairs received No blind interpretation All discrepant results reviewed Quality Score=1.5
from 500 practitioners by independent Ref. Std: 0
who chose to offer Conventional smear cytopathologist Blind: 0
ThinPrep along with and ThinPrep Verification: 0
conventional smear. thresholds of minor Most of HSIL and inconclusive Consecutive: 0
24% of women changes/HPV, LSIL, results on either ThinPrep or Spectrum: 0 Publication: 1
undergoing screening inconclusive conventional smear had Industry: .5
pap, and 35% of abnormality, HSIL, histologic confirmation
women needing referral cytologic reference
pap had both samples standard for test No test negatives verified, or
taken negatives. Histologic those with LSIL
reference standard for
those with HSIL or Relative True Positive rate
inconclusive results (ThinPrep)=178/158=1.13
on either conventional
smear or ThinPrep Relative False Positive
rate(ThinPrep)=37/33=1.12

39,408 conventional No blind interpretation Screening test: ThinPrep Quality Score=4
Paps, and 10,694 Prevalence(not specified)= Ref. Std: 2
ThinPrep specimens, Pap and ThinPrep 42/54=78% Blind: 0
collected concurrently specimen with Se (LSIL,NS)=40/42=95% Verification: 0
in two patient threshold of LSIL Sp (LSIL,NS)=7/12=58% Consecutive: 0
populations, described Spectrum: 0
only by age GS: histology with Screening test: Pap Publication: 1
threshold of "positive" Prevalence(NS)=67/89=75% Industry: 1
Se (LSIL,NS)=57/67=85%
Sp (LSIL,NS)=8/22=36%

C-39