AYAs wanted to feel appreciated, not judged.

AYAs valued having a peer navigator who could ‘walk with them’ by sharing empathetic experiences.

AYAs wanted to engage on topics other than HIV.

AYAs valued professional appearance.

Pretrial navigator training adapted to emphasise empathy and storytelling, and professional appearance during visits.

Peer navigator training included sexual reproductive health and relationship education modules with appropriate referral options.

AYAs go through trusted adults to access phones or communicate through in-person visits.

Modified study protocol to include AYAs with no mobile phone access in the control arm, which allowed them to be eligible for re-randomisation to one of the more intensive interventions including IP-Nav and CCT.

Confidentiality and privacy are paramount. AYAs preferred untraceable messages sent in coded language that avoid words like ‘HIV’, ‘viral load’

Concern about too frequent communication.

AYAs have personal preferences: each one would like engagement responsive to those preferences.

Modified study text messaging architecture to ensure messages sent were not traceable.

Harmonised message content with AYA preferences for coded language

Study protocol changed to reduce planned message frequency.

Peer navigators trained to assess and respond to individual-level messaging preferences (eg, option booster communication if desired).

AYAs wanted to increase their own skills in interacting with clinic-based, adult healthcare workers.

Enhanced the peer navigator training with role plays on self-management skills for capacity building AYAs to interact with clinics and providers to meet their needs.

Transition from adolescent to adult services is a significant area of concern and possible disruption.

Even though outside study scope, training on transition integrated into peer navigator intervention sessions.

Findings, including suggestions for peer-group transition and early awareness of transition shared with facility-level stakeholders.