Boyd 2019.

Study characteristics
Methods	RCT
Participants	Participants:mothers with postpartum depression (N = 24). Age range: parents ≥15 years of age with a child between 1 and 3 months old. Mean (SD) parents age in years was 26.4 (1.9) for the intervention group and 26.3 (1.8) for the control group. Recruitment: three urban paediatric primary care clinics in areas with high proportion of African American and recent immigrant patients. Recruited using advertisements at participating practices and through direct referrals from clinicians. Eligibility: at least 15 years old, have a child between 1 and 3 months old, have Internet access on a computer or a smartphone, and screen positive for depressive symptoms. Country: USA
Interventions	Intervention: social media only (n = 12) Parents Interacting with Infants social media intervention included eight weekly sessions with topics focused on depression psychoeducation and behavioural activation, infant temperament (personality), play, feeding, safety, sleep, parent–child interactions (laughter) and reading. Each weekly session included a narrated PowerPoint presentation, a video clip, questions for sharing experiences, an exercise, follow‐up questions and a summary. The content was uploaded on the Facebook secret user group in three instalments over the course of a week to allow participants the time to read the material, practise the exercises and post comments/ answers to the questions posed by the facilitator. Control: non‐social media control (n = 12) The in‐person intervention was held weekly over 8 weeks with the same weekly topics as the social media group. Each session began with an introduction to the topic, series of questions, dyadic activity, and a wrap‐up and discussion of homework. Weekly reminder calls and/or texts were sent about the upcoming session.
Outcomes	Feasibility, severity of depression symptoms, and perceptions of parental competence.
Equity	High‐income country, area with a large number of African American and recent immigrant families who live below the poverty line, mothers with postpartum depression.
Notes	Health behaviours: not applicable. Body function: not applicable. Psychological health: severity of depression symptoms was the only outcome reported for this category. Well‐being: not applicable. Mortality: not applicable. Adverse effects: not applicable. Secondary outcomes: perceptions of parental competence was the only outcome reported for this category.
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	"Quote "Randomisation sequences were generated using computer‐generated numbers prior to enrolment with assignments contained in sequential sealed opaque envelopes."
Allocation concealment (selection bias)	Low risk	"Quote "Randomisation sequences were generated using computer‐generated numbers prior to enrolment with assignments contained in sequential sealed opaque envelopes."
Blinding of personnel	High risk	Trial registration describes this 'open label'
Blinding of participants	High risk	Trial registration describes this 'open label'
Blinding of outcome assessment (detection bias) All outcomes	High risk	Trial registration describes this 'open label'
Incomplete outcome data (attrition bias) All outcomes	High risk	No attrition in intervention group but 30% of control group did not complete follow‐up.
Selective reporting (reporting bias)	Unclear risk	Outcomes listed in trial registration do not match those in the published report, e.g. acceptability, attendance at 6 or more sessions, feasibility, change in parenting stress, change in social support, change in parenting discipline.
Baseline characteristics similar	Unclear risk	"Quote "The in‐person group was more likely to have a family income of <$15,000 than the social media group" which may have influenced participant ability to adhere to the intervention.
Baseline outcome measurements similar	High risk	"Quote "Mothers in the in‐person group reported significantly higher levels of parenting competence than the mothers in the social media group. Our results showed that the mean total depression scores on the BDIII for social media group was within the moderate range of clinical severity; however the score for the in‐person group fell within the severe range."
Protection against contamination	Unclear risk	Secret Facebook group. No other details provided.