Accuracy of HPV self-collection compared to clinician-collected HPV testing and cytology: a meta-analysis

Meta-analyses show comparable clinical accuracy of PCR-based human papillomavirus (HPV) assays on self- compared with clinician-collected specimens. We extended these meta-analyses by comparing HPV testing on both samples with cytology from clinician-collected specimens. Studies published in PubMed, Embase, and Cochrane Library up to September 2024 were reviewed for inclusion. Studies had to report paired HPV self-collection, clinician-collected HPV testing, and cytology. Performance measures included sensitivity and specificity of atypical squamous cells of undetermined significance or worse and high-risk HPV positivity in self- and clinician-collected samples for cervical intraepithelial neoplasia 2 or worse. Accuracy data were pooled using the bivariate normal model for logit transforms of sensitivity and specificity. Sixteen full-text articles met inclusion criteria. Studies were heterogeneous and included referral populations, which may overestimate cytology sensitivity. Cytology was less sensitive in detecting cervical intraepithelial neoplasia 2 or worse (pooled sensitivity 87%; 95% confidence interval, 76–93) compared with HPV testing on self- (90%) and clinician-collected samples (93%) when restricting to PCR-based HPV assays. Overall, HPV self-collection sensitivity surpassed clinician-collected cytology testing, demonstrating that 3-year screening intervals, which are recommended for negative cytology results, are safe to use for negative HPV results from self-collected specimens. Longitudinal data of HPV testing on self- versus clinician-collected samples are lacking.