Self-collected vaginal swabs and urine tracked closely for high-risk HPV detection in South Asia
A multicountry comparison in Bangladesh, Nepal, and Pakistan found that two non-invasive self-sampling approaches produced broadly similar high-risk HPV detection results in sexually active girls and women aged 15 to 25 years. The analysis paired self-collected vaginal swabs with first-void urine from 753 participants and applied harmonized laboratory methods across sites, giving public health programs a practical look at how specimen choice may affect surveillance and screening strategies in resource-limited settings.
The work sits inside the larger GLOBE-HPV project, which is designed to estimate HPV prevalence and incidence across low- and middle-income countries in South Asia and Sub-Saharan Africa. For countries expanding HPV vaccination and considering community-based surveillance, the specimen question is not trivial. A test that can be self-collected, transported, and processed reliably without a pelvic examination can lower friction in outreach programs, especially for adolescents and young adults who may be reluctant to undergo clinician-based sampling.
Similar overall yield, with small but meaningful differences in what each specimen captured
The overall prevalence of 14 HR-HPV types was 8.6% in self-collected vaginal swabs and 7.2% in first-void urine. For the seven vaccine types included in current HPV vaccine formulations, detection was nearly aligned at 5.3% in vaginal swabs and 5.0% in urine. Rates for HPV 16 and 18 were also almost identical, at 2.3% in vaginal swabs and 2.4% in urine.
Those closely matched figures matter because HPV 16 and 18 remain the central oncogenic targets for prevention programs. If a surveillance method can recover those genotypes at nearly the same rate as vaginal sampling, public health teams may have more flexibility when designing field studies, school-based campaigns, or outreach programs in settings where acceptability and logistics can determine participation. The data suggest that first-void urine is not merely a convenience specimen; for key vaccine types, it can function as a credible alternative to vaginal sampling.
Still, the two specimen types were not interchangeable in every individual. Bi-directional type-specific discordance appeared, meaning each method sometimes detected types the other missed. This is a familiar feature of low-level HPV detection, where viral burden may be uneven across the lower genital tract and urinary stream. A specimen that performs well overall can still diverge at the margins when viral copy number is small or when shedding is patchy.
Concordance stayed high, but low viral load drove discordant results
Agreement between the two sampling methods remained strong overall, with kappa values above 0.7. That level of concordance supports the view that both approaches are capturing much of the same biologic signal. In practical terms, the same infected person is often likely to test positive regardless of whether the specimen comes from the vaginal epithelium or the first urine stream.
The discordant samples clustered around lower viral loads, reflected by higher real-time PCR cycle threshold values. This pattern is clinically plausible. When viral burden is low, small differences in specimen collection, cellular content, and assay sensitivity can change whether the result crosses the detection threshold. Urine may dilute target DNA more than vaginal swabs, while vaginal swabs may better sample infected epithelium directly. In women with early, transient, or low-copy infection, those biologic differences can shape the final readout.
Even so, the study did not show a broad separation between methods. For the main surveillance question—whether a population-level program can estimate HR-HPV burden using a non-invasive sample—the answer from these data is yes. The paired design strengthens that inference because each participant served as her own comparator, reducing the noise that comes from between-person differences in sexual exposure, age, and infection history.
Why this matters for vaccination-era surveillance
As HPV vaccination programs expand, surveillance systems need methods that are acceptable, scalable, and methodologically consistent across borders. Self-collection reduces the need for trained clinicians and exam rooms, and it can improve access for young people who might otherwise avoid screening. In this setting, both self-collected vaginal swabs and first-void urine offer a feasible route to monitor HPV circulation after vaccine rollout.
The distinction between the two specimens is likely to become operational rather than purely diagnostic. Vaginal swabs may remain the stronger choice when maximal analytic sensitivity is required for broader HPV typing, while urine may offer easier collection in community campaigns, schools, and hard-to-reach areas. Because the study used standardized DNA extraction and the Allplex HPV28 Detection PCR assay across sites, the comparison carries extra weight: the observed differences are less likely to reflect sloppy lab variation and more likely to reflect the biology of the specimen itself.
A public health program choosing between them may therefore ask a simple question: is the goal to capture every low-level infection, or to measure population burden with a method that people will actually complete? This study suggests that for many surveillance applications, either specimen can deliver a usable signal, and for the most clinically important vaccine-targeted genotypes the results were closely aligned.
Limitations remain in how far these findings can be generalized
The cohort was limited to sexually active girls and women aged 15 to 25 years, so the results cannot automatically be extended to older screening populations or to settings with different sexual behavior patterns. The analysis also focused on paired self-collected specimens under standardized conditions, which may not fully mirror the variability of routine field implementation. Sample handling, transport, and participant instruction can shift performance outside a research protocol.
The study also leaves open a practical question for program planners: whether the same level of concordance would hold when the goal shifts from surveillance to individual-level triage. Population estimates and clinical decision-making do not always demand the same threshold of sensitivity. A method that is strong enough for prevalence mapping may still need additional validation before it is used to direct care for an individual patient.
For now, the evidence supports a straightforward conclusion grounded in the data: in young sexually active women in Bangladesh, Nepal, and Pakistan, self-collected vaginal swabs and first-void urine produced similar HR-HPV results, with the clearest differences emerging when viral load was low and discordance became more likely.
- • PubMed PMID 42284270: https://pubmed.ncbi.nlm.nih.gov/42284270/