Increasing the Uptake of Mental Health by Ukrainian Refugees in Germany

How to cite

BEHAGHEL, Luc; Avdeenko, Alexandra; Ette, Andreas; Spieß, C.Katharina; Obminski, Artur; Hazard, Yagan; Rüsch, Nicolas; Moldavski, Alexander; Duflo, Esther. 2024. "Increasing the Uptake of Mental Health by Ukrainian Refugees in Germany." AEA RCT Registry. AEARCTR-0013354. https://doi.org/10.1257/rct.13354-1.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Luc BEHAGHEL* , Alexandra Avdeenko , Andreas Ette , C.Katharina Spieß , Artur Obminski , Yagan Hazard , Nicolas Rüsch , Alexander Moldavski , Esther Duflo

* Primary investigator

Registry ID
AEARCTR-0013354
DOI
https://doi.org/10.1257/rct.13354-1.1
First registered
Last updated
Status
Completed
Trial dates
2024-04-22 – 2025-10-15
Countries
Germany ()
Registry record
http://www.socialscienceregistry.org/trials/13354

Abstract

Poor mental health among displaced populations is a widespread problem. Even where mental health (MH) care is available, uptake is limited by fear of stigma, and shame as well as by informational and language barriers. This randomized controlled experiment evaluates an anti-stigma intervention at scale implemented through social media promoting mental health uptake in a vulnerable population. The intervention focuses on Ukrainian refugees in Germany, a large group of displaced individuals with extensive access to digital services, and in a context where both cost and language barriers to access mental health support services are low. The study aims to test the impact of alternative messaging and dissemination strategies to increase the uptake of formal MH care. More specifically, the evaluated intervention puts three elements together. First, an easy-to-use Telegram bot was built specifically for the project to recruit first-order study participants (“seeds”), randomize seeds into nine groups, engage users to watch a video and facilitate the dissemination of the bot.
Second, videos are displaced in the bot to encourage help-seeking behavior and refer to a hotline and thirdly, a hotline run by Ukrainian-speaking medical professionals provides an accessible gate of entry to available mental health care. Our research is primarily designed to measure which video and dissemination strategy leads to more engagement with the hotline. The study builds upon existing knowledge on effective and cost-effective interventions to increase MH care uptake and aims to further reduce knowledge gaps by testing alternative approaches in a common, real-life setting.

Experimental design

The overarching research goal is to provide evidence of the impact at the scale of an
anti-stigma and information campaign promoting mental health uptake in a vulnerable population. To do so, we conduct a nationwide randomized controlled trial in
social media directly, leveraging and combining various elements:

• role models to convey anti-stigma messages,
• peer-to-peer dissemination of messages seeded within social networks,
• behavioral levers to reduce the psychological and material costs to action.

To test the impact of alternative messaging and dissemination strategies this evaluated intervention puts three elements together. First, an easy-to-use Telegram bot was built specifically for the project to recruit seeds, randomize seeds into nine groups, engage users to watch a video, and facilitate the dissemination of the bot. Second, videos are displaced in the study Telegram bot to encourage help-seeking behavior and refer to a hotline and thirdly, a hotline run by Ukrainian-speaking medical professionals provides an accessible gate of entry to available mental health care German-wide.

Ukrainian refugees are invited to the study Telegram bot in two main ways: (i) seed users are directly recruited through physical channels (e.g. posters and flyers at language centers) and through virtual channels (e.g. pinned messages in existing regional Telegram groups of Ukrainian refugees); (ii) “higher order” users are recruited by seeds and other users of the bot that share the link to the study bot.

The campaign uses videos that provide information on the hotline and the main features of formal mental health care in Germany (full coverage by health insurance, confidentiality), as well as anti-stigma messages conveyed by role models. The videos are displayed on the Telegram bot built specifically for the project, which offers four follow-up actions: calling a hotline, requesting to be called, requesting to be reminded, and sharing options. The two videos are in Ukrainian language. They distinguish themselves by the initial sequence:
- The first video features celebrities (celebrity video) that invite speaking to others about emotional problems, and to seek help;
- The second features a Ukrainian woman who tells about her experience of receiving mental health care in Mannheim and inviting to seek help (relatable peer video).
The experimental hotline is served by three Ukrainian-speaking medical professionals based in Germany, some of whom have a refugee background. The main goal of the hotline is to offer assistance in navigating the German mental health system for Ukrainian refugees in Germany. The hotline builds on a preexisting hotline at a university hospital in Mannheim (ZI) and was expanded for the intervention to act as a nationwide hotline.

Design details

Randomization method
The sample consists of Ukrainian refugees in Germany who click on the link to the study Telegram bot (seed users and higher-order users). The recruitment will be through various channels to reach a sufficiently large number of Ukrainian refugees in Germany belonging to various subpopulations. The seeds will be individually randomized by the study bot into nine groups.
Randomization unit
Seed users will be individually randomized to one of eight treatment arms or an active control situation. The eight treatment arms are obtained from cross-randomizing the two video treatments, the two dissemination strategies, and two alternative action buttons (2 × 2 × 2 factorial design). They are tested against each other and against an active control situation where seeds are only given a short written description of the hotline. The video treatments include watching a video with celebrities’ messengers or a relatable peer. Regarding the dissemination, members of the ally treatment arm will be asked to become allies or ambassadors in the information campaign and receive additional support on how to share the link to the bot while members from the neutral dissemination treatment arm will be provided only with a “share” button but no further instruction. The last variation emerges by displaying either a “be called button” or a “reminded button” to lower the bar for action.
Clusters
-
Observations
We expect to recruit 2000 seeds. 2/3 of them will be recruited in the first weeks of the implementation phase and the last 1/3 will be recruited after 6 weeks.
Arms
The randomization is stratified by recruitment channels. The 2000 seeds are divided into 9 groups, whereby 9% of seed users are assigned to each treatment arm, and the remaining 28% are assigned to the active control group.
Primary outcomes
1.) Help-seeking behavior: calls to the hotline 2.) Social media shares of the videos 3.) Medium-term effects: change in mental health care uptake, attitudes on mental health, and perceived barriers to help-seeking for callers and Telegram users.

Cited by

No citation data available for this registration yet.

Keywords