Back

HPV self-sampling for cervical cancer screening during wartime: a pilot implementation study from a conflict-affected region of Ukraine

Kovalyov, O.; Kovalyov, K.; Deinichenko, O.; Dillner, J.; Arroyo Muhr, L. S.

2026-01-02 public and global health
10.64898/2026.01.01.26343310 medRxiv
Show abstract

Structured AbstractO_ST_ABSIntroductionC_ST_ABSCervical cancer screening is difficult to initiate and sustain during war due to disrupted health services, population displacement and limited clinical capacity. Evidence on whether human papillomavirus (HPV)-based screening using self-sampling can function as a screening pathway during armed conflict is unknown. This study evaluated the feasibility, acceptability and operational performance of HPV self-sampling for cervical cancer screening in a war-affected region of Ukraine. MethodsWomen aged 30-60 years living in Zaporizhzhia Oblast, Ukraine, were offered HPV testing using self-collected cervicovaginal samples distributed through outpatient clinics, community outreach activities and refugee centres. Samples were analysed locally using PCR-based HPV assays with external laboratory quality assurance. Women positive for HPV16/18 were referred directly for colposcopic evaluation. Primary outcomes included screening uptake, kit return rate, HPV positivity and follow-up completion among HPV16/18-positive women. Secondary outcomes included turnaround times, HPV positivity by age and sampling setting, and participant feedback. ResultsBetween April and December 2025, 593 women were offered HPV self-sampling; only four (0.67%) declined participation. Of 589 distributed kits, 572 samples were returned and analysed (attendance rate 96.46%, 572/596), with non-return occurring only in refugee centres. Overall, 52/572 women (9.1%) tested positive for targeted HPV types, including 22 (3.8%) positive for HPV16/18 and 30 (5.2%) positive for HPV31/33/35/45/52/58. All women testing positive for HPV16/18 were contacted and referred for specialist follow-up; among those completing follow-up, all were assessed within two months. Median turnaround time from self-sampling to results was 6.3 days. All recorded participant feedback was positive. ConclusionHPV self-sampling-based cervical cancer screening can be initiated and maintained during active war with high uptake, timely laboratory processing and effective referral of women at highest risk. Simplified, low-contact screening pathways may support cervical cancer prevention under wartime conditions. Trial registrationClinicalTrials.gov identifier: NCT07275333. What is already known on this topicO_LIHPV-based screening is the most effective method for cervical cancer prevention and is recommended globally. C_LIO_LIHPV self-sampling increases participation and acceptability in stable healthcare settings. C_LIO_LIWhether cervical cancer screening can be implemented during an active armed conflict is unknown. C_LI What this study addsO_LIDemonstrates that HPV self-sampling-based screening can be initiated and sustained during active war. C_LIO_LIShows high uptake, timely laboratory processing and effective referral of HPV16/18-positive women under severe operational constraints. C_LIO_LIIdentifies distribution models that minimise sample loss in highly mobile and displaced populations. C_LI How this study might affect research, practice or policyO_LISupports inclusion of HPV self-sampling in cervical cancer prevention strategies for conflict-affected settings. C_LIO_LIInforms programme design choices, favouring simplified, single-visit screening pathways when population mobility is high. C_LIO_LIProvides implementation evidence relevant for countries rebuilding screening programmes during or after military conflicts. C_LI

Published in BMJ Global Health (predicted rank #15) · training set

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.