Colposcopy performance in the new primary HPV screening in Australia: How to determine colposcopy competency?

Abstract

Aims

To assess colposcopic performance and determine indicators for competency within the new Australian primary human papillomavirus (HPV) cervical screening program.

Materials and methods

A retrospective observational study of 4542 women seen at The Royal Women's Hospital Colposcopy Clinic in Melbourne, from 1 December 2017 to 31 July 2020 after a higher-risk cervical screening test (CST) result.

Results

Histological CIN2+ was detected in 25.1% up to two years from first colposcopy visit (FCV). The majority (86.7%) of CIN2+ was detected early within the first six months of presentation. Biopsy rate overall was 96.1% with abnormal colposcopic impression. Of four colposcopists with a lower biopsy rate, only one was able to achieve this early detection rate. Biopsy was also taken in over 30% of cases with negative reflex cytology and normal colposcopy, with CIN2+ detected in 5.0% among positive HPV16/18 and 3.8% with non-16/18 HPV. Positive predictive value of high-grade colposcopic impression at FCV averaged 66.4% (range: 54.9–81.6% among our colposcopists) and is poorly correlated with early detection rate of CIN2+. Overall accuracy of colposcopy is 84.5% (range: 78.7–90.3%), buoyed by high true negative colposcopic predictions secondary to high rates of negative reflex cytology referral with the new screening algorithm and is also unlikely to be a useful colposcopy indicator.

Conclusions

Early detection rate of CIN2+ within the first six months of presentation is a useful measure of colposcopy competency and we would encourage our National Cancer Screening Register to explore this with the participating colposcopists.