Bigna 2015.

Study characteristics
Methods	Aim: to determine 1. if reminders sent by text message, phone call, or concomitant text and phone calls most increase the presence at medical appointments of HIV‐infected or HIV‐exposed children, and 2. which is the most efficient related to working time and financial cost. Study design: parallel individual RCT Recruitment: adult–child (carer–patient) pairs were recruited for HIV‐infected or exposed children aged < 15 years attending HIV care. Study duration: not reported Study dates: January 2013 to May 2013
Participants	Inclusion criteria: people aged ≥ 18 years accompanying an HIV‐infected or HIV‐exposed child aged < 15 years for HIV care Sample size (n): 242 (text message: 60; text message + call: 61; call: 60; control: 61) Age: children: age range up to 15 years, although we could not confirm that ≥ 70% of included children were ≤ 5 years we decided to include this study since the weighted mean age of children was 2.8 years. Carers (mean): text message + call group: 36.5 (SD 12.4) years; text message group: 41.4 (SD 12.8) years; call group: 50.5 (SD 13.2) years; control group: 42.6 (SD 12.6) Sex: carers – men: text message + call group: 20%; text message group: 13%; call group: 15%; control group: 13%; children – boys: text message + call group: 52%; text message group: 47%; call group: 47%; control group: 49% Country: Cameroon (lower middle‐income) Setting: 3 hospitals serving urban (Essos), semi‐urban (Kousséri), and rural (Goulfey) settings
Interventions	Intervention: appointment reminders via text message; voice call; or text message + voice call Content: appointment reminder Frequency and intensity: 1 call, or 1 text message then 1 call 2–3 days before appointment Control: standard care/no intervention. No reminder (usual practice), all participants attended standard HIV care appointments Co‐interventions: HIV care (to all participants)
Outcomes	1. Attendance at HIV medical appointments; 2. Direct costs; 3. Staff working time Outcome assessment time point: 2 days
Funding/declarations of interest	Funding: no external funding, done as part of an academic degree financed from the personal funds of the corresponding author Jean Joel R Bigna. Conflicts of interest: authors declare no competing interests
Notes	Trial ID: PACTR201304000528276
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Computer‐generated sequence. Quote: "Randomisation and allocation were done centrally with WinPepi version 11.25. Eligible participants (adult–child pairs) were randomly assigned in blocks of four and allocated (1:1:1:1) sequentially in the order of receipt of a randomisation code."
Allocation concealment (selection bias)	Low risk	Central allocation. Quote: "Randomisation and allocation were done centrally with WinPepi version 11.25. Eligible participants (adult–child pairs) were randomly assigned in blocks of four and allocated (1:1:1:1) sequentially in the order of receipt of a randomisation code."
Blinding of participants and personnel (performance bias) All outcomes	High risk	Personnel were blinded. Not possible to blind participants. Quote: "The treating physician, the medical administrative assistant responsible for contacting participants in the intervention groups 2, 3, and 4 via mobile phone … were all masked to group assignments."
Blinding of objective outcome assessment (detection bias)	Low risk	Outcome assessors and analysts were blinded. Quote: "the nurse (outcome assessor) responsible for recording the patient's presence or absence at the appointment, and the data analysts were all masked to group assignments."
Blinding of subjective outcome assessment (detection bias)	High risk	Self‐reported measures would have been at high risk of detection bias since the study was not blinded.
Incomplete outcome data (attrition bias) All outcomes	Low risk	All randomised participants, whether or not they received the intervention, were included in the ITT analysis.
Selective reporting (reporting bias)	Low risk	All outcomes listed in the protocol were fully reported.
Other bias	Unclear risk	Some baseline characteristics (ages of children and carers, carers' education, and time to appointment) were unevenly distributed between groups.