Eze 2015.

Study characteristics
Methods	Aim: 1. To provide evidence validating the need for development and deployment of automated client reminder‐recall systems for the Nigerian National Routine Immunisation Program and 2. to compare its projected cost with the cost of a health personnel‐based defaulter tracking system. Study design: parallel individual RCT Recruitment: carers were selected using multistage sampling method from 8 health facilities in Egor local government area of Edo State. Study duration: each child recruited into the study at its first immunisation session (BCG) was followed up for 18 weeks while those who were recruited at their second session (DPT1) were followed up for 12 weeks to allow for 4 extra weeks after the recommended dates of receipt of DPT3. Study dates: June 2010 to June 2011
Participants	Inclusion criteria: bringing child for routine immunisation for the first or second schedules Sample size (n): 1001 (intervention: 500; control: 501) Age (mean): carers: 29.35 (SD 5.3) years Sex: 895 (89%) mothers, and 10 classed as 'other' but sex not stated Country: Nigeria (lower middle‐income) Setting: facilities included 1 tertiary hospital, 2 primary health centres, and 5 privately owned health facilities all of which provided routine immunisation services.
Interventions	Intervention: reminder and recall text messages Authors also report that (quote), "All text messages were by internet‐based web‐to‐SMS (Bulk SMS) service and were tagged the name of client's health facility for easy recognition." Messages sent the day before an immunisation appointment. There were 4 immunisation time points for BCG, DPT1, DPT2, and DPT3. Content: short reminder text message, and recall messages to those who missed appointments Frequency and intensity: reminder messages sent 1 day before the appointment, the number of reminder messages received depended the time point at which the participant was recruited. Recall messages were sent to defaulters. The total number of these messages was not reported. Control: standard care/no intervention, details not reported Co‐interventions: none reported
Outcomes	1. Timeliness of receipt of DPT3; 2. Cost and cost‐effectiveness (compared to estimated cost of home visits) Outcomes reported but not included in the review: 1. Barriers to receiving text message reminders (outcome not eligible for inclusion) Outcome assessment time point: postintervention, exact follow‐up time not reported
Funding/declarations of interest	Funding: not reported Conflicts of interest: not reported
Notes	Trial ID: not reported
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Quote: "Participants recruited per facility were randomized into 2 equal groups: intervention and Control groups using the RANDOM. EXE function of the Programme for Epidemiologists (PEPI) version 4.0."
Allocation concealment (selection bias)	Unclear risk	Not reported.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Not possible to blind participants or personnel.
Blinding of objective outcome assessment (detection bias)	Low risk	Quote: "Data collectors could not tell if a client was in the intervention or control group."
Blinding of subjective outcome assessment (detection bias)	High risk	Unblinded carers interviewed using follow‐up questionnaires.
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Available‐case analysis with 9.6% (96 carers) attrition. Reasons for dropping out were provided and dropouts were balanced between groups.
Selective reporting (reporting bias)	Unclear risk	All expected outcomes and outcomes listed in the methods section were reported in the results, but no protocol or online trial record was identified to check against.
Other bias	High risk	Some participants switched groups after randomisation. Authors stated that, "Some respondents, who did not have mobile phones but were randomized into the Interventional group initially, were eventually matched for age and sex and swapped with persons who own mobile phones and were randomized into the Control group." Comment: in addition, carers provided answers to questions by recall; this could have led to recall bias in the information supplied.