Impact of a Classroom-based Sensitization Intervention on Demand for Mental Health Care Among Adolescents in India

Background and rationale:

Schools are recommended as strategically important platform for improving young people's access to mental health care, including in low and middle- income countries (Patton et al., 2016). Along with a limited supply of effective treatments, low awareness about mental health problems and their treatment often act as barriers to help seeking in schools. Interventions to increase mental health literacy have been shown to increase the demand for mental health care among young people in a number of studies conducted worldwide, including in low-resource settings. However, the evidence is not conclusive as many of the existing studies are associated with a high risk of bias (Wei et al., 2013; Wei et al., 2015; Kutcher et al., 2016).

Formative research in Government-run secondary schools in New Delhi, India indicated that mental health services were non-existent and adolescents had limited knowledge about mental health care. The investigators have therefore developed sensitization activities to promote referral to, and uptake of, a newly developed school counselling service. These sensitization activities have been designed for implementation at both whole school and classroom-levels, and are intended to be delivered by non-specialist counsellors. The additional resource investment in classroom-level sensitization activities may be justified if this component leads to significantly greater levels of demand. This trial therefore seeks to evaluate the additional impact of the classroom sensitization session on demand for school counselling in New Delhi.

This study is part of a larger research program called PRIDE (PRemIum for aDolEscents). The overall goal is to develop and evaluate a trans-diagnostic, stepped-care intervention targeting common mental disorders in school-going adolescents in India. The PRIDE intervention architecture involves two sequential psychological treatments of incremental intensity. The first step is a brief problem-solving intervention delivered by lay counsellors. Adolescents with enduring mental health problems are referred to a higher intensity trans-diagnostic treatment delivered by psychologists.

The trial described here is embedded within a host trial that is testing the effectiveness of the low-intensity problem-solving intervention. The embedded trial will evaluate the impact of the classroom sensitization activities over and above the school-level activities on referral rates generated for the host trial.

Study design and setting:

An embedded recruitment trial will apply a stepped-wedge, cluster (class)-randomized controlled design in 70 classes across six Government-run secondary schools in New Delhi.

Eligibility criteria: Please see section on eligibility.

Interventions: Please see section on Intervention arms.

Outcomes: Please see section on outcome measures.

Trial scheme:

The trial will be conducted over 12 weeks and recruit 70 classes. The referral outcomes for the students will be reported every four weeks across these classes. As each class has an average enrollment of 50, the study will follow around 3400 students for the entire trial duration of 12 weeks.

Each class will switch over from the control condition to the intervention condition at 4-week intervals (excluding holidays and exam breaks), over 2 steps. During the first 4-week period, only school-level sensitization activities will be implemented. In the next 4-week period (first step), 35 randomly selected classes (sequence 1) will receive the classroom sensitization intervention. In the final 4-week period (second step), the remaining 35 classes (sequence 2) will receive the classroom sensitization intervention. Schedules for sensitization in the allocated classes will be shared with the schools in advance to ensure access.

Power calculation:

We based our power calculation on a within-period comparison for a stepped wedge design (Thompson et al., 2018) using Stata package "clustersampsi". Based on pilot data, we anticipated referral rates of 5% and 15% on the control and intervention conditions respectively, with an intra-cluster correlation coefficient (ICC) of 0.124. Using these parameters, a sample size of 70 classes (average class size of 50 students) will have 92% power to detect a difference of 10 percentage points (treating the outcome as a binomial variable), at a significance level of 0.05.

Data collection:

The primary outcome (referral rate) will be collated from referral logs maintained by researchers in each school. Referral data will be aggregated over each 4-week period. Secondary outcomes pertaining to the eligibility and clinical characteristics of students referred to the host trial will be derived from the Strengths and Difficulties Questionnaire (SDQ; Goodman et al., 2000).

Statistical analysis:

The baseline characteristics of the participating 70 classes, including class size and gender composition, will be described and assessed for any systematic differences across the two sequences. The primary outcome will be analyzed using generalized estimating equations (GEE) with robust standard errors. GEE is a recommended method for analysis of stepped-wedge cluster randomized controlled trials, providing population-averaged effects of exposure across trial conditions (Barker et al., 2016). GEE allows for longitudinal data analysis without resorting to fully specified random effect models and can be applied to both continuous and categorical outcomes (Liang and Zeger, 1986). It provides both parameter estimates and standard errors that are corrected for clustering of data and are consistent despite misspecifications in the correlation structure. For this trial, the clustering of data will be specified at the class level. Analysis of the secondary and exploratory outcomes will also be undertaken using the GEE method. Sensitivity analysis will be conducted using a 'within-period comparison' of data from the second period only (Thompson et al., 2018). No interim analysis will be undertaken.

Process evaluation:

We will undertake descriptive analysis of quantitative process data in order to explore potential sources of variation in the implementation of intervention procedures.