Stigma barriers to information diffusion: an experiment with refugee mental health care-seeking

Last registered on August 22, 2023

Pre-Trial

Trial Information

General Information

Title
Stigma barriers to information diffusion: an experiment with refugee mental health care-seeking
RCT ID
AEARCTR-0010702
Initial registration date
December 27, 2022

Initial registration date is when the trial was registered.

It corresponds to when the registration was submitted to the Registry to be reviewed for publication.

First published
January 03, 2023, 5:27 PM EST

First published corresponds to when the trial was first made public on the Registry after being reviewed.

Last updated
August 22, 2023, 11:16 AM EDT

Last updated is the most recent time when changes to the trial's registration were published.

Locations

Region

Primary Investigator

Affiliation
Harvard University

Other Primary Investigator(s)

Additional Trial Information

Status
In development
Start date
2023-01-02
End date
2024-05-31
Secondary IDs
Prior work
This trial does not extend or rely on any prior RCTs.
Abstract
This study uses a randomized controlled trial to study the take-up of mental health services in the context of a "sender" who has private information about an available mental health service, and a "recipient" who may be in need and with whom the sender may share the information. The study is conducted with a sample of Syrian refugees living in Jordan. In this setting there is a large mental health burden, with roughly 50% of the adult population having symptoms aligned with clinical depression (Stillman et al. 2022). Senders are invited to share with their friends mental health awareness information advertising a free counseling helpline. The sender's choice to share is observed by the researcher, as is the recipient's decision of whether to take-up the helpline. I consider first what constrains sharing of information, and second how message recipients are impacted, both in their probability of using the helpline as well as their perceptions about their own fit for the service relative to the reputational costs of using it.

Amendment 8/22/23
This amendment adds a longer-term primary outcome (recipient's desire to be contacted directly by the helpline), and adds a follow-on randomized intervention (when the recipient is directly contacted by the helpline).

Citations
Stillman, Sarah, Rozo, Sandra V, Tamim, Abdulrazzak, Palmer, I Bailey, Smith, Emma and Miguel, Edward, (2022), The Syrian refugee life study: first glance, Oxford Review of Economic Policy, 38, issue 3, p. 625-653.
External Link(s)

Registration Citation

Citation
Smith, Emma. 2023. "Stigma barriers to information diffusion: an experiment with refugee mental health care-seeking." AEA RCT Registry. August 22. https://doi.org/10.1257/rct.10702-3.1
Experimental Details

Interventions

Intervention(s)
The intervention consists of a mental health awareness campaign, delivered over WhatsApp over the course of 8 days. The mental health awareness campaign includes information on identifying psychological distress, self-care tips, and information about a free helpline service. The campaign content was designed and piloted locally by the International Rescue Committee in Jordan.

Amendment 8/22/23
Female recipients from this study and recipients in the separate study AEARCTR-0011928 are together randomized into the second intervention, which is direct contact from the phone counseling helpline. The new intervention consists of being contacted by the helpline directly, with the participant’s prior consent. Which participants are randomly sent first to the helpline to be contacted, and therefore begin receiving counseling services first, is randomized. Delaying some participants’ treatment is necessary due to helpline capacity constraints, providing the rationale for randomization. When first offered the possibility of the helpline call participants are never guaranteed the helpline call, and are informed that even if they do receive the helpline call that, due to capacity constraints, is could be multiple months before they are contacted.
Intervention (Hidden)
Amendment 8/22/23
Approximately 600 total participants are expected in the experiment with roughly 100 new participants identified weekly, though this will depend on both participants’ interest in the helpline and the survey rate of the enumerators.
The helpline has capacity constraints that prevent it from admitting more than approximately 30 new beneficiaries per week. (This number is subject to change as the helpline learns through experience.)
Intervention Start Date
2023-01-09
Intervention End Date
2024-01-31

Primary Outcomes

Primary Outcomes (end points)
Sender follow-through
Take-up of helpline
Recipient's perceived mental health benefit relative to reputational cost
Recipient’s mental health stigma
Amendement 8/22/23:
Interest in helpline direct call
PHQ-9 score
Deperssion
GAD-2 score
Anxiety
Primary Outcomes (explanation)
Sender follow-through is defined two ways:
- A sender-level binary variable indicating whether the sender shared the content with at least 1 recipient.
- A recipient-level binary variable indicating if the sender sent at least 1 message to the recipient.
Take-up of helpline: Constructed two ways. First as the number of helpline caller phone numbers that match the recipient phone numbers in each arm, collected via the sender survey and recipient surveys.
Mental health benefit relative to reputational cost: The respondent's predicted benefit to self of using mental health services, minus the respondent's predicted reputational cost to self of using mental health services, with both component variables and the final variable normalized relative to the control group.
Recipient’s mental health stigma : Recipient’s willingness to share her experience publicly to be used for future IRC content. Normalized score of not willing at, willing when anonymous, to willingness when not anonymous.
Amendment 8/22/23
Interest in helpline direct call: As a new longer-term recipient outcome to the initial intervention, female recipients are contacted by phone directly, roughly 6 months after treatment, and offered the possibility of having the helpline contact them personally, if they are found to be eligible for the helpline phone counseling service. The indicator variable of whether individuals accept that offer is a primary outcome. Treatment heterogeneity in whether the individual was depressed at baseline is of particular interest.
Outcomes for the randomized direct helpline contact:
PHQ-9 score: Constructed using the 9-item Patient Health Questionnaire.
Depression: An indicator for whether the PHQ-9 score is 10 or larger, which follows the standard cutoff guidance for the PHQ-9.
GAD-2: Constructed using the 2-item Generalized Anxiety Disorder questionnaire.
Anxiety: An indicator for whether the GAD-2 score is 3 or larger, which follows the standard cutoff guidance.

Secondary Outcomes

Secondary Outcomes (end points)
Secondary Outcomes (explanation)

Experimental Design

Experimental Design
“Senders” will be invited to share mental health awareness content with their friends. Conditional on stating their willingness to do this, the senders are randomized to either a control group, which receives no information until after the study, or a treatment group. The treatment group senders receive the intervention information (the mental health awareness campaign) over WhatsApp over the course of 8 days. Each time they are requested to share the information with their friends, who have been identified prior to the randomization. Each treatment sender will be randomized to one of three different message framings.

The intended recipients of the messages form the recipient sample. They are surveyed 3 weeks after the intervention ends, and asked about mental health care-seeking and stigma attitudes.

Amendment 8/22/23
Participants that agree to be contacted by the helpline and are found eligible in a short screening consisting of the PHQ-9 and GAD-2 are then randomized on a roughly weekly basis to “Treatment Now” or “Treatment Later 1” or “Treatment Later 2”.
Experimental Design Details
Treated senders are randomized to one of three message framings, which appears in the first, introductory message that the senders is asked to share with their friends:
1. “The IRC is compensating me to share this with all of my close friends.”
2. “The IRC is compensating me to share this with friends who I think can benefit from the information.”
3. “I want to try to share this with all of my close friends.”

Amendment 8/22/23
Treatment Now: Phone numbers shared with the helpline to be contacted as soon as possible and may begin receiving phone counseling.
Treatment Later 1: Phone numbers shared with the helpline after all of Treatment Now has been shared, meaning approximately 6 weeks after the screening survey.
Treatment Later 2: Phone numbers shared with the helpline after Treatment Later has been contacted, meaning approximately 12 weeks after the screening survey.


The treatment assignment procedure is described below assuming it is conducted weekly; if due to field logistic constraints it is conducted every two weeks then the samples will be scaled accordingly
1. Treatment Now group: Draw a random subset of 30 people, stratifying on 1) whether she was a participant in the first intervention (AEARCTR-0011928), 2) whether she was in the treatment group in the first intervention, 3) tercile of distress (constructed using the sum of the PHQ-9 and GAD-2 scores).
2. Treatment Later 1: Draw a second random subset of 30 people or the total remaining sample if fewer than 30 participants remain.
3. Any remaining participants form the “Treatment Later 2” group.
Randomization Method
Randomization done in office by a computer.
Randomization Unit
Senders
Amendment 8/22/23: For the direct helpline intervention randomization is at the individual participant level.
Was the treatment clustered?
Yes

Experiment Characteristics

Sample size: planned number of clusters
900 senders
Amendment 8/22/23: 600 participants in the direct helpline contact experiment
Sample size: planned number of observations
3600 recipients and 900 senders Amendment 8/22/23: 600 participants in the direct helpline contact experiment
Sample size (or number of clusters) by treatment arms
225 senders control, 675 senders treatment. 225 senders per message framing in the treatment group.
Amendment 8/22/23: Direct helpline contact experiment: 180 in Treatment Now; 180 in Treatment Later 1; 240 in Treatment Later 2.
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
Sender follow-through (binary variable): 10.4 percentage point difference in follow-through assuming 50% follow-through in the benchmark framing (standard deviation .5). Follow-through measured at the recipient and sender levels. Take-up of helpline (binary variable): 5.0 percentage point difference in follow-through. Baseline mental health care-seeking in this population is 7% with a standard deviation of 25 percentage points. Measured at the recipient level. Amendment 8/22/23 Based on baseline levels of depression and anxiety in the eligible population the MDEs on the primary outcomes comparing Treatment Now to the pooled Treatment Later groups are: PHQ-9 score: 1 point change in the score. Depression: 6.8 percentage point change in rate of depression. GAD-2 score: 0.4 point change in the score. Anxiety: 10.6 percentage point change in rate of anxiety.
Supporting Documents and Materials

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information
IRB

Institutional Review Boards (IRBs)

IRB Name
Harvard University-Area Committee on the Use of Human Subjects
IRB Approval Date
2022-04-04
IRB Approval Number
IRB20-1023
Analysis Plan

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Post-Trial

Post Trial Information

Study Withdrawal

There is information in this trial unavailable to the public. Use the button below to request access.

Request Information

Intervention

Is the intervention completed?
No
Data Collection Complete
Data Publication

Data Publication

Is public data available?
No

Program Files

Program Files
Reports, Papers & Other Materials

Relevant Paper(s)

Reports & Other Materials