Abstract
OBJECTIVE: To undertake a comprehensive assessment of the strength of preferences among young people for attributes of emerging technologies for testing and treatment of asymptomatic chlamydia. DESIGN: Discrete choice experiment (DCE) with sequential mixed methods design. A staged approach to selection of attributes/levels included two literature reviews, focus groups with young people aged 16-24 years (n=21), experts' review (n=13) and narrative synthesis. Cognitive testing was undertaken to pilot and adapt the initial questionnaire. Online national panel was used for final DCE survey to maximise generalisability. Analysis of questionnaire responses used multinomial logit models and included validity checks. SETTING: England. PARTICIPANTS: 1230 young people aged 16-24 from a national online panel (completion rate 73%). OUTCOME MEASURES: ORs for service attributes in relation to reference levels. RESULTS: The strongest attribute influencing preferences was chlamydia test accuracy (OR 3.24, 95% CI 3.13 to 3.36), followed by time to result (OR 1.81, 95% CI 1.71 to 1.91). Respondents showed a preference for remote chlamydia testing options (self-testing, self-sampling and postal testing) over attendance at a testing location. For accessing treatment following a positive test result, there was a general preference for online (OR 1.21, 95% CI 1.15 to 1.28) versus traditional general practitioner (OR 1.18, 95% CI 1.12 to 1.24) or pharmacy (OR 1.15, 95% CI 1.10 to 1.22) over clinic services. For accessing a healthcare professional and receipt of antibiotics, there was little difference in preferences between options. CONCLUSIONS: Both test accuracy and very short intervals between testing and results were important factors for young people when deciding whether to undergo a routine test for asymptomatic chlamydia, with test accuracy being more important. These findings should assist technology developers, policymakers, commissioners and service providers to optimise technology adoption in service redesign, although use of an online panel may limit generalisability of findings to other populations.