Closing the Mental Health Gap in Jordan using Behavioral Design: The InforMH Project
Co-authored with Rachit Shah and Joy Georgi
Mental health is a critical issue that affects people worldwide, but it becomes even more crucial in emergency and refugee contexts. In Jordan, a country hosting the second largest number of refugees per capita and still recovering from the COVID-19 pandemic, a significant portion of the population is affected by lasting distress, including depression, anxiety, and PTSD. However, only a small fraction of people receive the proper care they need. Some reports suggest that 94% of people with mental health conditions in Jordan are not receiving treatment. Barriers to accessing mental health services include a lack of awareness surrounding the need for care, stigmas around mental health, and lack of trust in health services. As in many countries, mental health remains a taboo topic in Jordan, with significant cultural barriers to identifying challenges and seeking care.
To address these issues and bridge the treatment gap, the Airbel Impact Lab developed a program called “InforMH”, which we are calling “Yalla Nihki” locally in Arabic. While many programs seeking to expand mental health services are focused on increasing the supply of those services, InforMH is geared toward helping people overcome the obstacles keeping them from using the services in the first place. We are harnessing the power of behavioral science and human-centered design to help people access the help they need.
InforMH was co-designed with hundreds of community members in Jordan and all components of the program have been tested iteratively with them over the last 3 years. Below, we’ll walk you through our design and testing process, and two of the big questions we learned answers to along the way.
As an introduction to the program, here’s a short video:
What kind of messaging would encourage more people to seek out care?
We know some of the main reasons people do not seek out care is because they feel weak or ashamed of needing mental health support. In this project we wanted to learn what type of messaging might address, or overcome, this stigma.
In behavioral science, we often use A/B tests to compare two versions of something and learn which one might be more effective at encouraging people to change their behavior. In this case, we wanted A/B testing to help us answer, “What type of messaging is most likely to encourage people to engage with the InforMH program?”
These tests are often used in the tech world and social marketing. In our context, they can be used to improve program design, by generating evidence on what works and doesn’t work at impacting outcomes related to the uptake of mental health services.
For our A/B test, we compared two versions of messaging that had both received positive feedback from people during prototyping. Both messaging framings are also backed by behavioral science evidence:
- Version A’s messaging focussed on endorsement of the program by respected experts, professionals, and showing how the program was backed by data. Behavioral science literature shows that humans are heavily swayed by the influence of authority figures, especially experts like doctors and professionals.
- Version B’s version of messaging, focussed on endorsement by other people like them, in Jordan. We used this because behavioral science evidence also suggests that we humans are more likely to do something if we see that others like us are also doing it.
We reached over 400,000 people with this test on Facebook. Of these people, around 3,700 clicked through to our website which included mental health psycho-education content, self-care recommendations and a button to call a hotline. These messaging options were not just shown to people on Facebook, but were also reflected on two different versions of the website (A and B), each with the different framing woven throughout.
Results
While it was very close, Option B — which focussed on peer endorsement — was ultimately more successful. It led 14% more unique visitors to the InforMH website and those who went to Website B spent 33% more time on it. Finally, Option B also led to 30% more clicks on a button which called a mental health hotline, however, this was not statistically significant given the number of smaller people who did this action. These results showed us that this messaging provided a direct change, increasing help-seeking from people who might otherwise not have sought care.
Since Option B performed better at encouraging behavior change among the wider Jordanian population, we integrated this messaging, promoting peer endorsement by other relatable people in Jordan, throughout many areas of the program.
How did people experience the whole program?
The InforMH program reaches people both online through social media and in person at locations they already go to for help, such as pharmacies. The program first invites people to take an online questionnaire that assesses their mental health needs. Based on their results, people are provided with tailored recommendations to pursue either formal, informal, or self-care treatments for mental health needs. The program takes an ‘ecosystem’ approach, meaning that people are connected to existing care, throughout the mental health system, that’s right for them. This could be informal care, such as a community support group, or formal care, through a trained mental health counselor. InforMH then uses behavioral-science informed messaging and reminders on WhatsApp, to help participants follow through with their suggested care regimen, such as framing or goal setting.
Once we learned what messaging and framing would most encourage people to seek out support, it was time to launch a short “test run” of the entire program for a month. Until this point, we had tested the components of the project individually, but we needed to test the overall user experience. In this 2022 month-long test, we took the components developed over three years with community input –from the self-assessment to the behavioral messages — and delivered it through a WhatsApp bot and tested it together as a package. We aimed to learn how people experienced the entire journey, from when they found out about InforMH, took the self-assessment and received tailored recommendations, to when they seek out the form of care advised to them.
We designed this 31-day journey for users, which was then delivered through an automated chat-bot. Delivering through a chatbot allowed us to easily reach people daily where the majority already are, on Whatsapp, without needing them to download a separate app or find a site. The bot offered almost daily messages, encouragement, reminders to take action, psycho-education information and more.
We selected 30 participants and 21 of them actually started the journey. The remaining 9 people didn’t continue either due to some technical issues they were experiencing, because they were busy or because they ran out of internet data.
Learnings
Most users had very positive feedback on the program, with the majority of people using words to describe it as “excellent”, “helpful” or “beneficial”. Specifically, most people felt this way because the program was:
- Personal and customized: Several users expressed that the program made them feel like someone cared about them and their wellbeing, sometimes for the first time. One user said: “I felt that the program cares about me individually”.
- Informative: Users mentioned that they’re benefiting from it and that the information is helpful. One user mentioned: “I now understand more about anxiety…. the questions [self-assessment questionnaire] helped me understand myself and the situations better”, while another already started seeing the benefits: “Even people around me started noticing the difference, my habits were changing for the better. I used to sleep for 2 hours only.”
- Accessible: Users loved being able to access the content from the comfort of their own home.
- Helpful: Users appreciated the constant reminders and accountability to take action for themselves.
However, there were also some very important lessons we learned about how we need to change the program to meet more client needs. Some small technical glitches created barriers to the experience and led to some participants stopping their journey for days at a time, or even altogether. For example, some participants were not sure about how to start the chatbot conversation. The program team has been working with the platform developers to remove these barriers, and make it a more seamless and easy experience to engage in.
Secondly, participants still experienced uncertainty and fears around confidentiality. In other words, even though we anonymised their identities on the bot, some participants were concerned that people might find out about them seeking mental health support once they sought the care services we recommended. Due to societal stigma, they were concerned about what others would think if they knew. This led to lower than anticipated calls to formal and informal care services. For the next iteration of the program, we plan to add more messages speaking to these fears. We also plan to share a video of what calling the hotline or our care-provider partners might look like, to assuage some fears or uncertainties for those who feel nervous. Additionally, even though some participants expressed concerns about confidentiality, we found that many participants felt more comfortable sharing their mental health experiences with a chatbot than with a real person, as it created a safe and non-judgmental space for open conversation.
Finally, some participants said that in the 31-day challenge they wanted to see more tangible signs of their progress. In other words, “How do we know we’re making progress?” In future phases, the program would be delivered over months, not weeks, which would allow participants to see real changes in their lives. We will not only include more behavioral messaging to help them see the effects of their efforts, but we will also provide participants with interactive and reflective ‘check points’ so they can also report on their progress. We will also offer participants testimonial videos from others who have benefited from the program, so they can see how others like them are gaining real help from InforMH. This will be complemented by allowing people to take the self-assessment questionnaire during or after the program, so they can see the changes they have made in terms of their mental health needs.
Next steps
This year, these learnings will be integrated into an improved program that will be delivered as a 3 month long “pilot study”, with hundreds of people across the country to learn about how the program works in real world conditions. These next phases of this project will set us on a path to launch our intervention nationwide — with the potential to benefit 1.65 million people in Jordan with unmet mental health needs. Countries like Jordan are still affected by the long-lasting effects of the pandemic and thus breakthrough innovations are required to counter the potentially devastating impacts of the mental health treatment gap on individuals, families and communities. If we can close the mental health treatment gap in Jordan, then InforMH has the potential to scale globally — confronting stigma around mental health across the many countries affected by crises.
