Are government or non-governmental organization staff more effective in delivering frontline mental health services? Evidence from a cluster randomized controlled trial

The proposed study will test the effectiveness of two implementation models for delivering group Problem Management + (gPM+). gPM+ is an intervention developed by the World Health Organization in 2013 to address common mental health problems such as depression, stress, and anxiety. It involves problem management (PM) plus (+) selected behavioral strategies to address both psychological issues (e.g., stress, fear, feelings of helplessness) and, where possible, practical challenges. The first delivery model relies on health extension workers (HEWs), public employees already under government supervision who are trained and equipped by an NGO to deliver gPM+. The second entails local facilitators recruited for a short-term assignment and compensated hourly (i.e., not on a full-time basis) by an NGO.

Following the training and practicum period, each individual delivering gPM+ (HEW or LF) is assigned two gPM+ groups including up to eight individuals each, in order to deliver five 90-120 minute group sessions weekly over five weeks. The number of groups is consistent for all facilitators (whether HEWs or LFs), and the total anticipated workload for each facilitator is around 6—8 hours weekly for a period of five weeks (three to four hours per group). As the provision of mental health support is a responsibility for HEWs and thus within their scope, HEWs do not receive any additional compensation for engaging in this new activity. LFs are receiving a monthly payment for three months, compensating them for time committed to preparation, group facilitation, and follow-up during this period.

2024-05-06

2024-07-01

Depressive symptoms as measured by the PHQ-9,
Stress as measured by the Perceived Stress Scale (PSS-10),
Post-traumatic stress as measured by the abbreviated PCL-C instrument,
Anxiety as measured by the General Anxiety Disorder 7 (GAD-7)
Child development as measured using the Strengths and Difficulties Questionnaire

The evaluation is conducted within four districts or woredas (Samre, Tsaeda Emba, Keletawelaelo, and Enderta) of Tigray and uses a cluster randomized control trial (cRCT) design. Randomization was conducted at the village level (the sub-administrative unit below the kebele), stratified by kebele. The total number of villages included in the sample is 222 across 61 kebeles.
Within each kebele, villages were randomized to receive one of the three experimental arms described below;
• Control: no intervention
• Treatment 1: gPM+ delivered by HEWs
• Treatment 2: gPM+ delivered by LFs
Within T2, there was a further sub-randomization in which villages were randomly assigned to either a male or female LF.

Each HEW (who are all women) was assigned to facilitate one gPM+ group for men and one gPM+ group for women, and thus, eight men and eight women were recruited for gPM+ in each village served by a HEW. Male LFs were assigned to facilitate gender-matched gPM+ groups, while female LFs were assigned to facilitate one male gPM+ group and one female gPM+ group. Thus, in villages served by male LFs, 16 men were recruited to participate in gPM+, while in villages served by female LFs, eight women and eight men were recruited.

In both HEW and LF arms, individuals recruited to participated in gPM+ were PSNP clients (either direct support or public works) of working-age (between 18 and 64 years old), who were heads of household or the spouse of the household head, and who exhibited signs of moderate to moderately severe depression (PHQ-9 score between 10 and 19, inclusive).

Randomization was done in an office using STATA.

Unit of randomization was the village.

Yes

222 villages

3,552 individuals

70 control villages; 64 HEW villages, 88 LF villages (44 male and 44 female)

Are government or non-governmental organization staff more effective in delivering frontline mental health services? Evidence from a cluster randomized controlled trial