Evaluation of Pre-Analytical Variables for Human Papillomavirus Primary Screening from Self-Collected Vaginal Swabs

Human papillomavirus (HPV) primary screening is an effective approach to assessing cervical cancer risk. Self-collected vaginal swabs can expand testing access, but the data de fi ning analytical performance criteria necessary for adoption of self-collected specimens are limited, especially for those occurring outside the clinic, where the swab remains dry during transport. Here, we evaluated the performance of self-collected vaginal swabs for HPV detection using the Cobas 6800. There was insigni fi cant variability between swabs self-collected by the same individual ( n = 15 participants collecting 5 swabs per participant), measured by ampli fi cation of HPV and human 13-globin control DNA. Comparison of self-collected vaginal swab and provider-collected cervical samples ( n = 144 pairs) proved highly concordant for HPV detection (total agreement = 90.3%; positive percentage agreement = 84.2%). There was no relationship between the number of dry storage days and amplifi cation of HPV ( n = 68; range, 4 to 41 days). Exposure of self-collected dry swabs to extreme summer and winter temperatures did not affect testing outcomes. A second internal control (RNase P) demonstrated that lack of ampli fi cation for 13-globin from self-collected specimens was consistent with poor, but not absent, cellularity. These data suggest that self-collected vaginal samples enable accurate clinical HPV testing, and that extended ambient dry storage or exposure to extreme temperatures does not in fl uence HPV detection. Furthermore, lack of 13-globin ampli fi cation in HPV-negative samples accurately identi fi ed participants who required recollection.