Collecting national sexual health survey and biological data: experiences of Natsal-4

Great Britain’s fourth National Survey of Sexual Attitudes and Lifestyles (Natsal-4) was designed to provide high quality representative data on the sexual and reproductive health, attitudes and lifestyles of people aged 16-59 years. The study included the option for all consenting participants to self-collect either three vaginal swab samples or one urine sample, for testing of sexually transmitted infections and the microbiome. Participants were also asked to give consent to link their data to health, education and/or administrative records.

Originally scheduled for 2020 as a face to face probability survey, Natsal 4 required substantial adaptation due to the COVID 19 pandemic. Implemented in 2022-2024, the final design comprised three data collection arms (1) an address based probability sample using in person interviewing with a telephone option; (2) a probability panel sample completing either a one hour telephone interview or a shorter online survey; and (3) a non probability online panel completing a 20 minute survey. Biosampling and linkage consent were included across all arms.

Eligibility for biosampling included all probability-sample participants and a subset of non probability participants (aged 18–49, plus those aged 50–59 in specific ethnic minority, trans/gender diverse, or men who have sex with men groups). Eligible consenting participants self collected either three vaginal swabs (cisgender women) or a first void urine sample (cisgender women declining swabs, cisgender men, and trans/gender diverse participants). Samples were tested for Chlamydia trachomatis, Neisseria gonorrhoeae, human papillomavirus (HPV), Trichomonas vaginalis, and Mycoplasma genitalium, and for genital microbiome analysis, with no return of individual results.

This presentation discusses the feasibility, challenges, and successes of implementing biosample collection within a multi-mode, multi-sample national sexual health survey. Methodological implications are relevant for both cross-sectional and longitudinal surveys.