Vaginal self-sampling without preservative for human papillomavirus testing shows good sensitivity.

Background: Several strategies have been used to reach non-attending women in organized cervical-cancer-screening programs, with varying success. Self-sampling (SS) for HPV is effective for increasing coverage in screening programs, but requires expensive commercial sampling kits.

Objective: We aimed to evaluate if vaginal SS, without commercial preservatives was adequate for HPV testing.

Study Design: Women with abnormal cervical smears as determined from the organized screening program were invited to a colposcopy clinic. The 121 women were asked to insert a cotton swab into the vagina and rotate it, put the cotton swab into a sterile cryotube, break the upper part of the stick and put the cap on. Thereafter, the gynaecologist collected a liquid based cytology (LBC) sample. The presence of HPV-types in SS and LBC samples was analysed with PCR and luminex-based typing.

Results: High-risk-HPV (hr-HPV) DNA was found in 65 of the tested 108 SS (60%; 95% CI 0.50-0.69), whereas LBC found hr-HPV in 64/108 samples (59%; 95% CI 0.49-0.69). The agreement between sampling with SS and LBC was good, kappa value 0.67 (95% CI; 0.53-0.81). The sensitivity for SS with hr-HPV to find HSIL was 77% (95% CI; 62-91%), specificity 47% (95% CI; 35-59%) [corrected] and the sensitivity for LBC with hr-HPV to find HSIL was 79% (95% CI 66-93%), specificity 50% (95% CI; 38-62%). [corrected]

Conclusions: This new vaginal self-sampling method detects hr-HPV-infections with similar sensitivity as a cervical smear taken by a gynaecologist. This self-sampling method is cost-effective and well tolerated, and the kit is suitable for regular mail transport.