Evaluation of the sensitivity of cervicography in a consecutive colposcopic series

Tumori. 1992 Jun 30;78(3):211-3. doi: 10.1177/030089169207800313.

Abstract

Cervicography was performed in 606 women referred for colposcopy. Cervigrams were blindly reviewed by two independent readers. The positivity rate at cervicography was high (operator A = 50%, B = 58.8%). The sensitivity for papillomavirus infection (HPV)/cervical intraepithelial neoplasia I (CIN I) (n = 141) was 79.4% for operator A and 80.8% for operator B. The sensitivity for CIN II or more severe lesions (n = 22) was 95.2% and 90.5% for operators A and B, respectively. The positive predictive value for HPV/CIN I or CIN II, or more severe lesions was 36.9% and 6.9% for operator A and 32.1% and 5.3% for operator B, respectively. Interobserver variability was acceptable (kappa = 0.62). Cervicography suspected 27 HPV/CIN I, 1 CIN II and 1 CIN III which showed no cytologic abnormalities. This study confirms that cervicography has a good sensitivity for cervical lesions, but it is based on a selected series, not representative of a screening condition. The combination of cervicography and cytology in screening is presently under evaluation in a prospective study of screened women.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Carcinoma in Situ / diagnostic imaging*
  • Cervix Uteri / diagnostic imaging*
  • Female
  • Humans
  • Middle Aged
  • Observer Variation
  • Papillomaviridae
  • Radiography
  • Sensitivity and Specificity
  • Tumor Virus Infections / diagnostic imaging
  • Uterine Cervical Neoplasms / diagnostic imaging*