Digital delivery of Behavioural Activation to overcome depression and facilitate social and economic transitions of adolescents in LMICs (DoBAT)

The DoBAt study will be an individual randomised control trial in which participants are randomly allocated to each of the intervention and control arms. All participants in the intervention and control arms will be given a low-end smartphone (which they can keep at the end of the study).

Participants in the intervention arm will receive Behavioural Activation therapy delivered via a smartphone app implemented over 10 weeks. The app consists of six tailored modules that contain Behavioural Activation’s core principles integrated into a gamified format. The app will be supplemented by brief phone calls (15 mins per module) from Trained Peer Mentors.

Participants in the control group will get access to an app that contains short video clips from WildEarth-SafariLive, a locally produced wildlife documentary series.

2021-11-25

2023-01-31

(1) Feasibility and acceptability of the intervention
(2) Initial efficacy (direction and magnitude) of any effects of the intervention on depressive symptoms

(1) Feasibility and acceptability of the intervention. A mixed-methods approach will be adopted to establish feasibility and acceptability.
Feasibility will be assessed by collecting data on the following:
• Recruitment (enrolment rate of eligible participants) and retention in the trial at the end of the intervention period (11 weeks).
• Feasibility of testing procedures and data collection methods, including assessment of completion rates.
• Treatment adherence rates, where adherence is defined as having opened at least 4 out of 6 of the app modules and as having completed 3 out of 6 phone calls with the peer mentor (excluding the introductory call). We will complement the treatment adherence rates with engagement metrics collected via the app (number of times participants logged into the app, number of modules opened and completed, total time spent on the app, number of weekly activities set to do, and number of times the participant completed the weekly activities).

Acceptability of the intervention and study procedures will be assessed via:
• An acceptability questionnaire conducted at the end of intervention assessment (week 11) with all participants. The questionnaire consists of three measures: Acceptability of Intervention Measure (AIM), Intervention Appropriateness Measure (IAM), and Feasibility of Intervention Measure (FIM). Each measure consists of 4 items. The total score ranges from 1 to 5 and is calculated by averaging response scores to the response categories. We will calculate an average score for each of the measures. We will inquire about the acceptability of the app using these three measures (to both the intervention and control groups) and the acceptability of the peer mentor programme (to the intervention group only).
• In-depth interviews with a subsample of participants. We estimate that interviews with 20 participants will provide a sufficient range of experiences and perspectives to reach data saturation. This subsample will contain participants from the treatment and control arms and will be stratified based on high versus low app engagement.

Fidelity of delivery of the intervention will be assessed by collecting data on adherence and competence of Trained Peer Mentors.
• Adherence of the peer mentors is defined as the number of sessions that meet at least 90% of the criteria for adherence according to the training protocol. Independent raters will listen to a random sample of recordings of Peer Mentors’ phone calls with participants and assess them against the training protocol.
• Competence of the peer mentors will be expressed as a percentage based on their Competency Assessment Test. This test will include a written test and observation of skills through role-playing to assess Knowledge, Attitudes and Practices. Tests will be scored by the Trial Psychologist using a pre-determined scoring system.


(2) The initial efficacy of the intervention will be assessed by looking at differences between treatment and control group in the scores of the 9-item Patient Health Questionnaire Adolescent version 9-item (PHQ-A) at the end of intervention assessment (week 11). Outcomes will be compared between intervention and control groups using a linear regression model and adjusting for covariates (PHQ score measured at baseline, sex, age, depression severity, and household asset index). As a secondary analysis, we will make use of the repeated measurements of the PHQ-A throughout the trial (up to 6 times per individual) to evaluate the treatment effects over time.

Pilot locally adapted outcome measures of mental health, cognition, risky behaviours, socioeconomic measures, and collect descriptive data to inform the development of a further larger trial.

Mental health outcomes: General Anxiety Disorder (GAD-7), Conor-Davidson Resilience Scale (CD-RISC), Brief Rumination Response Scale (RRS), Emotional wellbeing (WEMWBS), Behavioural Activation for Depression Scale (BADS).

Cognition outcomes: emotion recognition, working memory, task switching, risk-taking, non-verbal reasoning

Risky behaviours: risky sexual behaviours, substance use, delinquency, gambling

Economic measures: time preference, risk preference, loss aversion. In addition, we will measure a range of socioeconomic outcomes, including measures of time use, human capital investment, and spending.

This pilot study is a two-arm single-blind individual-level randomised control trial. Enrolled participants will be individually randomised into intervention or control arms of the study (100 participants each).

Participants will be randomly assigned to the intervention or control arm with a 1:1 allocation using a computerised minimisation algorithm, balanced by sex (male or female) and severity of depressive symptoms (<10 or ≥ 10 on the PHQ-A). The minimisation algorithm was generated by the Centre for Healthcare Randomised Trials (CHaRT) at the University of Aberdeen.
Recruitment will be staggered over 6 months.

Individual level randomization

No

200 individuals.

200 individuals.

100 individuals in treatment arm, 100 individuals in control arm.

Statistical power was calculated to detect differences between two independent groups, in a two-sided test with an α of 0.05 and a power of 1–β=0.80, for an effect size of 0.45. This effect size is based on findings from previous studies investigating digital psychological interventions, with effect size (Cohen’s d) ranging from 0.24 to 0.57. Given these findings, we chose a small-to-medium effect size of 0.45 to account for the limited number of studies conducted with our target population and in our study setting. We allowed for 25% attrition, based on a previous study with adolescents in the study setting. Given these assumptions, we aimed to recruit 200 participants at baseline. We calculated the required sample size using G*Power software, version 3.1.9.3.