Clinical Validation of a Vaginal Cervical Cancer Screening Self-Collection Method for At-Home Use: A Nonrandomized Clinical Trial.

Importance: One-quarter of US women who are at risk for cervical cancer delay screening. Self-collected (SC) cervical screening was recently US Food and Drug Administration (FDA)-approved in the US for use in a health care setting only; an at-home SC option is crucial to address clinic-related barriers to screening. Objective: To clinically validate the use of an SC device that was designed for optimal at-home performance, safety, ease-of-use, and dry storage and transport. Design, Setting, and Participants: This nonrandomized clinical trial used a prospective method comparison study design. Participants aged 25 to 65 years were recruited from 16 clinical sites in the US including community and academic practices from November 20, 2023, to April 5, 2024. Data analysis was conducted from April to August 2024. Intervention: Eligible participants collected a sample with the SC method, followed by a clinician-collected (CC) sample. The SC sample was eluted into PreservCyt at the laboratory and both samples were tested on an FDA-approved high risk human papillomavirus (hrHPV) test approved for primary screening. Participants were followed up for safety and completed usability and screening preference surveys. Main Outcome and Measures: The primary outcome measures were positive percentage agreement (PPA) and negative percentage agreement for detection of hrHPV between the SC and CC samples. Other study measures included clinical sensitivity for high grade cervical dysplasia and usability. Results: Of 609 screening-eligible participants, 599 (262 aged 30-39 years [43.7%]; 583 identified as female [97.3%]) had paired SC-CC samples, of which 582 had valid paired samples included in the end point analysis. Among the 582 evaluable paired samples, the PPA between SC compared with paired CC samples for detection of hrHPV was 95.2% (95% CI, 92.1%-97.1%; 278 of 292 participants). The absolute clinical sensitivity for detection of high-grade cervical dysplasia was 95.8% (95% CI, 86.0%-98.8%; 46 of 48 participants), equivalent to the CC (relative sensitivity, 1.00). Nearly all participants (555 of 601 participants [92.3%]) reported that the device instructions were easy or very easy to understand and also that they would choose SC if they knew the results were comparable to CC results (560 of 602 participants [93.0%]). Conclusions and Relevance: In this nonrandomized clinical trial, SC samples collected with the device showed equivalent clinical sensitivity and exceeded the PPA end point for cervical screening. This SC method was found to be easy to use and to be a preferred option with high clinical performance intended for use in an at-home setting. Trial Registration: ClinicalTrials.gov Identifier: NCT06120205.