Abstract
Cervical cancer remains a major public health challenge, particularly in low- and middle-income countries. The primary cause of cervical cancer is high-risk human papillomavirus (HPV), and screening using physician-collected samples is complicated by stigma, inconvenience, and access. There are non-invasive alternatives to the physician-collected samples, including self-sampling methods such as first-void urine and vaginal swabs. This systematic review and meta-analysis evaluated and compared the diagnostic accuracy of vaginal and urine self-sample methods for detecting high-risk HPV. PubMed, Scopus, Web of Science, and the Cochrane Library were searched for studies published between January 2015 and October 2025. Bivariate random-effects models and HSROC models were used to estimate pooled sensitivity and specificity results compared with clinician-collected samples for CIN2+. Meta-regression assessed sources of heterogeneity. Twenty-two studies involving over 9000 participants were included. Vaginal self-sampling showed a pooled sensitivity of 91.3% and a specificity of 86.9%, while urine self-sampling showed 86.9% sensitivity and 79.5% specificity. Vaginal swabs demonstrated higher sensitivity in head-to-head comparisons. DNA-based PCR assays showed higher sensitivity than mRNA-based tests, and room-temperature storage decreased urine sample sensitivity. Both methods are effective for high-risk HPV detection. Vaginal self-sampling showed superior performance, while urine self-sampling remains a valuable non-invasive option for under-screened populations.