Structured CE was a new practice for most sites involved in the STREAM clinical trial. At all but one Stage 2 site, a new CE coordinating mechanism (in all cases, a CAB) was established around the time the trial began at the site. This meant there were no pre-agreed roles and responsibilities for key stakeholders related to CE. And, clearly defining and documenting the roles and responsibilities for CE in STREAM was not always prioritized at sites prior to initiation.
Despite this, STREAM CABs and study teams worked hard to build understanding of how their knowledge and skills were complementary and could be employed to improve trial results. In most cases, this meant involving CAB members as trusted partners for community outreach and (at some sites) participant support (financial and/or psychosocial). For CAB members, study teams invested in training CAB members to maximize their ability to participate as full partners throughout the clinical research cycle.
Despite the overwhelmingly positive experience of STREAM, there were a few areas of misunderstanding that arose from ill-defined roles. One example relates to contact between trial participants and CAB members. Although the benefits of peer support for MDR-TB patients are well-accepted, there is no consensus regarding the role (if any) CAB members should play in providing psychosocial support to trial participants. The confidentiality of research participants in trials is paramount, and therefore some sites elected not to support CAB member/ participant contact. On the other hand, CAB members were often drawn from NGOs and CBOs whose role in other contexts is to support people with TB and they were therefore expecting to play the same role in the trial. At study sites where this issue was not clearly addressed and agreed by stakeholders, diverging expectations had to be resolved.
A second example relates to the role of CAB members in recruitment. Trial sites are often under pressure to meet recruitment targets agreed with the trial sponsor and, in that context, it can be tempting to rely on CAB members (who often have deep roots in the community) to boost recruitment. On the other hand, the GPPs emphasize that the objectives of CE and recruitment activities are different, and that CAB members should not participate in participant recruitment. Differences in the expectations of STREAM study teams and CABs regarding their roles in recruitment required further engagement and discussion, although this improved over time.