Colposcopically obtained images of vulvar, vaginal, and cervical epithelium for assessment of safety of intravaginal agents among normal adolescent females: comparison with in vivo exam and interobserver agreement.

OBJECTIVES

This study was designed to assess identification of epithelial abnormalities of both in vivo examination as compared with colposcopically obtained images and interobserver assessment of the same images of the lower genital tract in healthy women.

MATERIALS AND

METHODS

Ninety women between the ages of 14 and 21 years were recruited for a phase II trial of a vaginal Lactobacillus crispatus capsule. All women underwent a baseline and 1-week colposcopic examination. Multiple genital tract areas were evaluated for abnormalities and photographed. The original examiner and two experienced colposcopists reevaluated all images masked to previous interpretations. Agreement was evaluated using kappa statistics.

RESULTS

The representative kappa statistics for direct observation vs photographic interpretation for the vulva, vagina, and cervix are: 37%, -2%, and -4%, respectively. The kappa statistics comparing the three observers ranged from 1% to 39%. CONCLUSIONS.: There is poor agreement between in vivo exams and photographic interpretation, and interobserver assessments of lower genital tract photographs.