Evaluation of Sample Pooling Strategies for Point‐of‐Care HPV DNA Testing Using the STANDARD™ M10 Platform
Ahmad Asyraf Senian, Vaenessa Noni, Abigail Rembui Jerip, Andy Cheong Shin Bong, Dzuriaty Siri‐ee, Cheng Siang TanABSTRACT
Human papillomavirus (HPV) DNA testing is the recommended primary screening method for cervical cancer prevention. However, implementation in resource‐limited and geographically challenging settings is constrained by cost, logistics, and time‐sensitive screen‐and‐treat workflow. Sample pooling may improve testing efficiency, but evidence for its feasibility on point‐of‐care molecular platforms remains limited. We evaluated the analytical performance and feasibility of sample pooling for high‐risk HPV (hrHPV) detection using the STANDARD™ M10 HPV point‐of‐care assay. Archived self‐collected vaginal swab samples from a rural cervical cancer screening programme, Programme SUARA, in Sarawak, Malaysia, were used, including four hrHPV‐positive samples spanning a range of cycle threshold (Ct) values and seven hrHPV‐negative samples. Pooling strategies of 2‐in‐1, 4‐in‐1, and 8‐in‐1 were assessed using a Dorfman two‐stage testing approach. Qualitative agreement, sensitivity, Ct shifts, and concordance with individual testing were analyzed. All hrHPV‐positive samples were consistently detected across individual and pooled formats, with no false‐negative results observed. Sensitivity remained 100% for all pooling strategies, with perfect agreement between pooled and individual testing ( κ = 1.0). Pooling resulted in modest Ct shifts, with mean increases of +1.0, +1.7, and +1.8 cycles for 2‐in‐1, 4‐in‐1, and 8‐in‐1 pools, respectively, without compromising qualitative detection, including samples with low viral loads. Sample pooling for hrHPV DNA testing using the STANDARD™ M10 platform is analytically reliable and operationally feasible. These findings support pooled testing as an evidence‐based strategy to improve throughput, reduce resource consumption, and enable same‐day screen‐and‐treat workflows in resource‐limited and remote settings.