The design and use of ACASI to elicit sexual behaviour in the PHASE survey – the Gambian experience

Abstract Collecting reliable sexual behaviour data remains a major challenge in conservative, low-literacy contexts such as The Gambia, where stigma, privacy concerns, and social desirability bias influence responses. The Prevalence of Human Papillomavirus, Antimicrobial resistance, and Sexually transmitted infection Estimation (PHASE) survey sought to address this challenge by introducing Audio Computer-Assisted Self-Interviewing (ACASI) for the first time in The Gambia. This paper documents the design, development, and piloting of the ACASI tool, highlighting key lessons for similar settings in sub-Saharan Africa. A Community Advisory Board (CAB) was central to the tool’s development, ensuring cultural sensitivity, acceptability, and linguistic accuracy. CAB members guided the drafting and adaptation of sexual behaviour questions, advised on culturally relevant response categories, and reviewed translations into four local languages (Mandinka, Serahule, Fula, Wolof). They also oversaw the design of non-threatening, familiar images to accompany response options and ensured accurate, context-appropriate audio recordings. Technical implementation was achieved using the SurveyCTO platform, allowing offline use, encryption, and integration of multilingual audio and images. Privacy safeguards were embedded, including screen codes replacing text to prevent disclosure in shared settings and restricted staff access to individual responses. Pilot testing revealed critical operational considerations. Many women lacked prior experience with digital tools and required extra support during the introductory phase. Adjustments included re-recording unclear questions, reordering items to reduce priming effects, and randomising response lists to enhance validity. Additional measures, such as repeating select questions at different points, improved internal consistency checks. Despite these challenges, participants demonstrated increasing confidence with the tool, and CAB feedback confirmed that ACASI was more acceptable for sensitive questions than face-to-face interviewing. Our experience demonstrates that ACASI is a feasible and culturally appropriate method to collect sexual behaviour data in The Gambia, even within low-literacy, conservative populations. The use of familiar images, validated translations, and community engagement enhanced usability, while technical refinements addressed privacy and confidentiality concerns. Key challenges related to computer literacy and dialect variation were addressed through training, piloting, and iterative refinement. This process illustrates that by prioritizing community involvement, technical adaptation, and cultural sensitivity, ACASI offers a promising framework to enhance the acceptability and potentially the accuracy of sexual behaviour reporting in large-scale epidemiological surveys. These lessons can be applied to similar sub-Saharan African contexts seeking to generate high-quality data to inform HPV (human papillomavirus), STI (sexually transmitted infection), and sexual health research.