McCann, Jainnee

Objectives: We evaluated agreement between human papillomavirus (HPV) tests performed on self-collected vaginal and clinician-collected cervical specimens to advance HPV self-collection as a cervical cancer screening option in our health care system. Methods: This diagnostic accuracy study was conducted at an academic medical center. Participants were 25–65 years old with upcoming colposcopy appointments. One cohort (n = 79) self-collected a vaginal specimen in the clinic immediately before a clinician-collected cervical specimen. The second cohort (n = 115) self-collected a vaginal specimen at home, mailed it in, and had a clinician-collected cervical specimen in the clinic. Paired specimens were evaluated for total (TPA), positive (PPA), and negative (NPA) percentage agreement. Extreme temperature stability studies were also conducted. Results: Compared with clinician-collected cervical specimens, in-clinic self-collected vaginal specimens showed a TPA, PPA, and NPA of 98%, 100%, and 97% for HPV16; 99%, 100%, and 99% for HPV18; 86%, 94%, and 79% for high-risk other (HRO) HPV. The TPA, PPA, and NPA for at-home self-collected vaginal specimens compared with clinician-collected cervical specimens were 94%, 79%, and 96% for HPV16; 95%, 43%, and 98% for HPV18; 89%, 80%, and 93% for HRO. Vaginal specimens showed prolonged stability and were unaffected by extreme temperatures. Conclusions: In general, in-clinic and at-home self-collected vaginal HPV specimens for HPV screening demonstrate high agreement with clinician-collected cervical specimens, with the exception of low PPA in the at-home self-collected vaginal HPV18 subgroup. Our findings support the use of HPV self-collection as a practical option to expand access to cervical cancer screening.