A systematic review and meta-analysis of diagnostic accuracy of urine self-sampling for human papillomavirus detection in cervical cancer screening

ObjectiveThis review assessed the diagnostic accuracy of urine self-sampling for human papillomavirus (HPV) detection compared with clinician-collected cervical and vaginal samples.MethodsA systematic search of PubMed, and Embase (January 2013-October 2023) identified full-text English studies evaluating urine-based HPV testing conducted in clinical settings and compared with provider-collected samples. Data were extracted using a standardised form, with sensitivity and specificity as primary outcomes. Risk of bias was assessed using QUADAS-2 and QUADAS-C tools. The review followed PRISMA-P guidelines and was registered in PROSPERO (CRD42024511154). A random-effects meta-analysis summarised diagnostic accuracy.ResultsOf 1657 citations, 13 studies met the inclusion criteria. Urine sampling showed lower sensitivity for detecting high-risk HPV and high-risk HPV-associated high-grade squamous intraepithelial lesions (HSIL) than vaginal samples, although specificities were comparable (87% for urine vs. 88% for vaginal samples). Sensitivity for detecting high-risk HPV-HSIL was 75% for urine, compared with 92% for vaginal and 95% for cervical samples. Cervical samples demonstrated the highest diagnostic accuracy, followed by vaginal samples, while urine samples performed the lowest on the summary receiver operating characteristic (ROC) curve.ConclusionUrine self-sampling is a non-invasive and acceptable option but demonstrates reduced sensitivity compared with vaginal and cervical sampling. Despite lower accuracy, its convenience may support screening among women who might otherwise not participate.