Objectives: Trials within Cohorts (TwiCs) is a promising design to make randomized trials more efficient. Cohort participants are asked for consent to be randomized into future low-risk interventions tested within the cohort. To enable TwiCs in the Swiss HIV Cohort Study, we added this "randomization consent" to the protocol and approached cohort participants subsequently for written consent. This study describes the TwiCs implementation process.

Study design and setting: We used a mixed methods design to evaluate the implementation process. We used cohort data to characterize participants accepting and declining randomization consent. We conducted a cross-sectional survey with cohort physicians to gather opinions and experiences regarding the TwiCs design. We did semistructured interviews with involved stakeholders (physicians, research personnel, participants, and ethics committee members) to get insights about attitudes, barriers, and facilitators implementing the randomization consent. In addition, we performed observations in cohort visits where the randomization consent was offered.

Results: Between July 2024 and July 2025, among 5297 cohort participants approached, 3067 (57.9%) accepted and 734 (13.8%) declined the randomization consent. In 1496 (28.2%) cases the decision was postponed to the next visit. Male sex, younger age, higher education, being consulted by a steady physician for at least three visits, and shorter cohort participation time showed higher acceptance rates. Interviewed participants cited fear of additional effort and a lack of interest in research as reasons for declining consent. The overall perception of TwiCs among cohort physicians and research personnel was positive. They recognized the potential to simplify the conduct of trials, especially to test low-risk interventions. Ethical concerns on the TwiCs consent procedure were rare. However, an explicit randomization consent was considered necessary by members of ethical committees while several physicians and participants felt positive about randomizing without explicit consent. The roll-out of the randomization consent was facilitated by well-trained, motivated personnel, and seamless integration into clinical routine. Main barriers for physicians in the consenting process were language barriers, participant difficulty understanding the concept, and time constraints due to tight consultation schedules.

Conclusion: The implementation of the TwiCs design, including the roll-out of a randomization consent in an existing, large-scale cohort, is feasible. The acceptance rate among participants was high.