Comparative efficacy, feasibility and acceptability of HPV DNA testing on first-void urine versus self-collected vaginal samples: a real-world study in a resource-limited setting

Introduction Self-sampling for cervical cancer screening is a promising strategy to improve coverage and reduce strain on health systems. First-void urine (FVU) has emerged as a non-invasive alternative to self-collected vaginal samples for detecting high-risk human papillomavirus (HPV). However, most supporting evidence is from colposcopy clinic settings and not from screening settings. This community-based study in Eastern India aimed to (1) assess agreement between HPV testing on FVU and vaginal samples, (2) compare their accuracy in detecting cervical intraepithelial neoplasia grade 2 (CIN2+) lesions and (3) evaluate feasibility and acceptability in a resource-limited setting. Methods At multiple rural clinics, 2500 women aged 30–60 years provided both FVU and self-collected vaginal samples. All samples were tested centrally using the Cobas4800 assay; discordant results were retested with the Allplex HR-HPV assay. HPV-positive women were referred for colposcopy and biopsy. Acceptability and preference were assessed through structured surveys with 1500 women and focus group discussions with health workers. These health workers counselled women at the clinics to provide both self-collected samples. Results All participants provided satisfactory samples, except for two FVU samples, which did not yield satisfactory test results on repeated analysis. Agreement between FVU and vaginal samples for HPV detection was 99.0% (κ=0.90), with high concordance for HPV 16/18 (κ=0.90). CIN2+ detection rates were 6.0 and 7.2 per 1000 women screened with FVU and vaginal samples, respectively (p=0.6), with no statistically significant difference in sensitivity. Among surveyed women, 98.1% preferred urine sampling. Health workers favoured both self-sampling methods over speculum-based clinician collection. Conclusion HPV testing using FVU demonstrates high agreement with vaginal self-sampling, comparable accuracy in detecting CIN2+ lesions and greater acceptability among women. This method is feasible and well-suited for cervical cancer screening in resource-limited settings.