Obstetrics & Gynecology · Medical Literature
Effect of Self-Sampling HPV DNA Testing on Cervical Cancer Screening in Singapore's Primary Care: Pragmatic Randomized Controlled Trial.
Abstract
In Singapore, cervical cancer screening rates remain suboptimal, with only clinician sampling included in the national screening program. Self-sampling human papillomavirus (HPV) DNA testing might overcome barriers to clinician-based screening and improve uptake. We assessed the effect of offering self-sampling alongside clinician sampling on screening uptake and high-risk HPV detection compared with clinician sampling alone. This pragmatic, 2-arm randomized controlled trial with Zelen design was conducted across NHG Polyclinics, NHG Health in Singapore. Women aged 30-69 years due for cervical cancer screening were randomly assigned (1:1) to intervention (offered selection of clinician sampling followed by self-sampling) or usual care (clinician sampling alone). The primary outcome was high-risk HPV DNA detection, and the secondary outcome was screening uptake. We performed subgroup analysis to examine both outcomes by prior screening history. Multivariable analysis identified factors associated with screening choices. From August 2024 to February 2025, 650 women were randomized, and 640 were included in the analysis (320 per arm). Significantly more high-risk HPV was detected in the intervention arm (3.1% vs 0.3%; absolute difference 2.8%; 95% CI, 0.8-4.8;P < .001), driven by greater screening uptake (56.6% vs 42.8%; absolute difference 13.8%; 95% CI, 6.0-21.4; P < .001), particularly among women with regular screening history and those with no prior screening. Among screened participants in the intervention arm, 56.4% chose self-sampling. Lower educational attainment was associated with lower odds of choosing self-sampling over declining screening. Offering self-sampling HPV DNA testing alongside clinician sampling significantly increased screening uptake and high-risk HPV detection rates. However, uptake remained lower among women with less education, highlighting the need for targeted engagement strategies.Abstract also available in: عربي (Arabic); Deutsch (German); Español (Spanish); Francais (French); हिन्दी (Hindi); Indonesia (Indonesian); (Chinese); (Japanese); Portugues (Portuguese).
Clinical relevance
Cervical cancer screening uptake remains suboptimal in Singapore, where the national program relies exclusively on clinician-based sampling — a model that perpetuates access barriers related to discomfort, privacy concerns, and healthcare avoidance. This pragmatic RCT directly addresses whether adding self-sampling HPV DNA testing as an option within primary care polyclinics can meaningfully increase screening participation among women aged 30–69 years, including the critically underserved never-screened subgroup. The findings are immediately relevant to primary care and gynecology practitioners in settings considering expansion or redesign of cervical cancer screening programs.
Key findings
1. Screening uptake was significantly higher in the intervention arm (self-sampling offered alongside clinician sampling) versus usual care (56.6% vs. 42.8%; absolute difference 13.8%; 95% CI, 6.0–21.4; P < .001). 2. High-risk HPV DNA detection rates were significantly greater in the intervention arm compared with the usual care arm (3.1% vs. 0.3%; absolute difference 2.8%; 95% CI, 0.8–4.8; P < .001). 3. Among participants in the intervention arm who underwent screening, the majority — 56.4% — actively chose self-sampling over clinician sampling, demonstrating strong patient preference for the self-directed modality. 4. The uptake benefit was observed across subgroups, including both women with a history of regular prior screening and those with no prior screening history, suggesting broad generalizability of the intervention effect. 5. Lower educational attainment was independently associated with reduced odds of choosing self-sampling over declining screening altogether on multivariable analysis, identifying a key equity gap. 6. The trial enrolled 650 women (640 analyzed, 320 per arm) using a Zelen randomization design across NHG Polyclinics from August 2024 to February 2025; no quantitative subgroup effect sizes or adjusted odds ratios were reported in the available abstract.
Practice implications
Primary care and OB/GYN clinicians should consider integrating self-sampling HPV DNA testing as a co-offered option alongside clinician sampling for women aged 30–69 years due for cervical cancer screening, as this strategy significantly increases both overall screening uptake and high-risk HPV detection without replacing the standard pathway. Patient selection should be inclusive — benefits were observed in both previously screened and never-screened women — but targeted outreach and culturally tailored education are essential for women with lower educational attainment, who demonstrated lower odds of selecting self-sampling and remain at heightened risk of continued non-participation. Implementation should include structured counseling on self-sampling technique, quality assurance protocols to ensure specimen adequacy, and clear follow-up pathways for any high-risk HPV-positive result irrespective of sampling method chosen.