Smartphone- and chat-based psychological treatment for adolescents with depression. A controlled trial
In adolescence the cumulative probability for depression rise to 20% and major depression becomes the primary cause of disability-adjusted life years worldwide. The disorder predicts a range of negative outcomes. Resources are insufficient. Moreover, many young individuals with depression do not seek professional help. Social stigma is the main reason why this population is underrepresented in primary care. Instead youth express willingness and need for online contact with health services. Particularly chat-services is requested. This reflects young people s use of internet in general: mobile applications such as instant messaging (chat) and text messaging are by far the most used media, often preferred over meeting face-to-face. To offer chat-based psychological treatment via smartphone is in line with younger patients preferences, and addresses the pressing question of how to make psychological treatment more available. We have developed a treatment comprised of 1) chat sessions and 2) reading material based on cognitive behavioural therapy. We propose to evaluate this treatment in a randomised controlled study to see if it can be an effective and feasible treatment for adolescent depression. If the treatment is effective, the project makes an important contribution to the research field of eMental health as well as to the overall initiative of improving mental health in youth.
Final report
The purpose of the project was to evaluate a new internet-based treatment program for adolescents with depression. Part of the novel aspects were that we integrated a real-time chat function and improved the functionality and appearance of the intervention. The project has followed the timeline and has generated more studies than we originally planned. We have also been able to support international research collaborations via a postdoc and the treatment program has been used in another study. It will also be used in the United States. As we found that the treatment approach worked we conclude that the second aim of the project which concerned feasibility, reduced therapist time, and equal adherence as in face-to-face treatments was achieved.
Several people were involved in the studies including persons who were not payed by the grant. This included students, Phd students, and psychiatrists. However the grant was used for postdoc, webmaster, researchers and costs in association with conducting three controlled trials, a qualitative study and open access publishing fees.
We started with a feasibility study which we later developed into a controlled study involving 70 adolescents (15 – 19 years old) with depression or depressive symptoms (Topooco et al., 2018). They were recruited from the general public but also via health care facilities and schools. We also used social media as a way to reach participants. The included participants were randomised to internet-based cognitive behaviour therapy (iCBT) or attention control. The treatment was delivered over 8 weeks and consisted of eight skill-based modules and eight weekly chat sessions. Modules targeted behavioural and cognitive factors. Techniques included psychoeducation, behavioural activation, cognitive restructuring, affect regulation, anxiety management, and relapse prevention. Modules comprised reading material corresponding to 6–10 book pages, educational videos, fictional patient stories, interactive tasks and homework. It was possible to access the intervention via computer or smartphone, but in a secure manner. All studies were approved by the ethics committee and we developed procedures for handling the participants in terms of security and recommendations about care contacts. When running the study we received comments that the chat sessions were too short (30 min) and decided to prolong them in the next study (Topooco et al., 2019). This study was similar in many ways but we updated the program based on comments from participants and also extended the chat sessions. We also reported one year follow-up data. The study included 70 participants as we had to exclude many (56%), given either few symptoms or more severe problems. We had initially planned to pool the results from the two trials but given very similar results (which means sufficient power) and the fact that the studies differed we decided to report them as two separate trials. As an additional interest in the trials we investigated knowledge gains and not only symptoms (Berg et al., 2019). We also did a qualitative study with 10 participants from the second study (Berg, Malmqvist et al., 2020). As we found improvements in knowledge and also questioned if the chat sessions were needed we decided to continue the project with a new trial. Based on our experiences from the first two studies we focused more broadly on mixed anxiety and depression in the same age group (Berg, Rozental et al., 2020). We did not include a control group but instead ran a factorial 2 x 2 design trial with 120 adolescents. We tested if chat sessions were associated with more improvements and also if knowledge could be boosted by improving our treatment from an educational point of view (Berg, Rozental et al., 2020). This part of the project was not mentioned in the application but emerged as a consequence of the findings in the first studies.
Both studies focused on depression showed moderate to large effects compared to the control groups on measures of depression and anxiety. This adds to the evidence-base regarding internet interventions for adolescents and also suggests that chat sessions are important. We also found that knowledge improved and that the treatment and therapists were appreciated by the participants. The factorial design trial could confirm the good results with large effects but also that knowledge is important as participants who received learning support improved more. However we did not find that the chat sessions were associated with better outcomes.
To summarize the most important results in three points:
1. Guided internet treatment with chat sessions lead to reduced symptoms of depression in adolescents aged 15-19 years.
2. Knowledge about treatment and depression increase following treatment.
3. Even if they are appreciated it may be that scheduled chat sessions are not needed and that the self-help program with email support is sufficient.
In relation to the international research front our findings stand out as being slightly better but also the fact that we are the only research group that has focused on the educational aspects of ICBT and not only symptom reduction. We are also early in starting to improve treatments by incorporating learning support. The technological aspects of our web solution should also be mentioned as the chat function is embedded in the treatment platform called iterapi. This development was implemented via the grant and also involves video chat. We however noticed that in our trials the adolescents preferred text chat over video, but this needs to be investigated further. Overall we provided treatment for 260 adolescents and more than 400 were screened for participation and advice in case of exclusion. The project has generated much data and there are many reports to be written as we have mainly focused on outcomes and not process research.
The project developed well and has generated many new questions. As stated the focus on learning support and the factorial design trial was directly derived from the project as we noted that comorbidity between anxiety and depression was very common. But perhaps most important is a finding that we have moved on with namely that many of the adolescents express problems with low self-esteem. We have mentioned this problem in our treatments but now decided to move on and develop a specific treatment for this problem. During the grant period the PI also was able to get funding for a project on young immigrants. This is a separate project but the experiences from the RJ project have been very important when adapting and translating interventions into Arabic and Dari/Farsi languages. Finally colleagues in another research group have used our ICBT program as a comparison treatment against psychodynamic internet treatment.
As mentioned a collaboration with researchers in the United States will hopefully generate a dissemination of our treatment into English. Overall our research involves collaborations with many researchers for example in Poland, Germany, Lithuania, and United Kingdom, but the specific focus on adolescents is so far in the United States via postdoc Naira Topocoo.
With regards to dissemination the results have been presented at conferences and meetings. The Covid-19 situation has had negative consequences for this part even if we have participated in online meetings. An important aspect of dissemination is the publishing of peer-reviewed papers in open access journals. All papers derived from the project have been published open access. With regards to groups outside the scientific community we have lectured for students and teachers but again this part has been hampered by the pandemic. The developed treatments are not yet part of regular health care but colleagues in Stockholm are working with implementation and we hope our results will be helpful in that process. Internet interventions researchers across the world meet at conferences and also national conferences. Unfortunately the most recent meeting in which we had planned to present our results was cancelled due to the pandemic. In conclusion, we managed to complete our studies before the pandemic but the dissemination has been affected by the current situation.
Several people were involved in the studies including persons who were not payed by the grant. This included students, Phd students, and psychiatrists. However the grant was used for postdoc, webmaster, researchers and costs in association with conducting three controlled trials, a qualitative study and open access publishing fees.
We started with a feasibility study which we later developed into a controlled study involving 70 adolescents (15 – 19 years old) with depression or depressive symptoms (Topooco et al., 2018). They were recruited from the general public but also via health care facilities and schools. We also used social media as a way to reach participants. The included participants were randomised to internet-based cognitive behaviour therapy (iCBT) or attention control. The treatment was delivered over 8 weeks and consisted of eight skill-based modules and eight weekly chat sessions. Modules targeted behavioural and cognitive factors. Techniques included psychoeducation, behavioural activation, cognitive restructuring, affect regulation, anxiety management, and relapse prevention. Modules comprised reading material corresponding to 6–10 book pages, educational videos, fictional patient stories, interactive tasks and homework. It was possible to access the intervention via computer or smartphone, but in a secure manner. All studies were approved by the ethics committee and we developed procedures for handling the participants in terms of security and recommendations about care contacts. When running the study we received comments that the chat sessions were too short (30 min) and decided to prolong them in the next study (Topooco et al., 2019). This study was similar in many ways but we updated the program based on comments from participants and also extended the chat sessions. We also reported one year follow-up data. The study included 70 participants as we had to exclude many (56%), given either few symptoms or more severe problems. We had initially planned to pool the results from the two trials but given very similar results (which means sufficient power) and the fact that the studies differed we decided to report them as two separate trials. As an additional interest in the trials we investigated knowledge gains and not only symptoms (Berg et al., 2019). We also did a qualitative study with 10 participants from the second study (Berg, Malmqvist et al., 2020). As we found improvements in knowledge and also questioned if the chat sessions were needed we decided to continue the project with a new trial. Based on our experiences from the first two studies we focused more broadly on mixed anxiety and depression in the same age group (Berg, Rozental et al., 2020). We did not include a control group but instead ran a factorial 2 x 2 design trial with 120 adolescents. We tested if chat sessions were associated with more improvements and also if knowledge could be boosted by improving our treatment from an educational point of view (Berg, Rozental et al., 2020). This part of the project was not mentioned in the application but emerged as a consequence of the findings in the first studies.
Both studies focused on depression showed moderate to large effects compared to the control groups on measures of depression and anxiety. This adds to the evidence-base regarding internet interventions for adolescents and also suggests that chat sessions are important. We also found that knowledge improved and that the treatment and therapists were appreciated by the participants. The factorial design trial could confirm the good results with large effects but also that knowledge is important as participants who received learning support improved more. However we did not find that the chat sessions were associated with better outcomes.
To summarize the most important results in three points:
1. Guided internet treatment with chat sessions lead to reduced symptoms of depression in adolescents aged 15-19 years.
2. Knowledge about treatment and depression increase following treatment.
3. Even if they are appreciated it may be that scheduled chat sessions are not needed and that the self-help program with email support is sufficient.
In relation to the international research front our findings stand out as being slightly better but also the fact that we are the only research group that has focused on the educational aspects of ICBT and not only symptom reduction. We are also early in starting to improve treatments by incorporating learning support. The technological aspects of our web solution should also be mentioned as the chat function is embedded in the treatment platform called iterapi. This development was implemented via the grant and also involves video chat. We however noticed that in our trials the adolescents preferred text chat over video, but this needs to be investigated further. Overall we provided treatment for 260 adolescents and more than 400 were screened for participation and advice in case of exclusion. The project has generated much data and there are many reports to be written as we have mainly focused on outcomes and not process research.
The project developed well and has generated many new questions. As stated the focus on learning support and the factorial design trial was directly derived from the project as we noted that comorbidity between anxiety and depression was very common. But perhaps most important is a finding that we have moved on with namely that many of the adolescents express problems with low self-esteem. We have mentioned this problem in our treatments but now decided to move on and develop a specific treatment for this problem. During the grant period the PI also was able to get funding for a project on young immigrants. This is a separate project but the experiences from the RJ project have been very important when adapting and translating interventions into Arabic and Dari/Farsi languages. Finally colleagues in another research group have used our ICBT program as a comparison treatment against psychodynamic internet treatment.
As mentioned a collaboration with researchers in the United States will hopefully generate a dissemination of our treatment into English. Overall our research involves collaborations with many researchers for example in Poland, Germany, Lithuania, and United Kingdom, but the specific focus on adolescents is so far in the United States via postdoc Naira Topocoo.
With regards to dissemination the results have been presented at conferences and meetings. The Covid-19 situation has had negative consequences for this part even if we have participated in online meetings. An important aspect of dissemination is the publishing of peer-reviewed papers in open access journals. All papers derived from the project have been published open access. With regards to groups outside the scientific community we have lectured for students and teachers but again this part has been hampered by the pandemic. The developed treatments are not yet part of regular health care but colleagues in Stockholm are working with implementation and we hope our results will be helpful in that process. Internet interventions researchers across the world meet at conferences and also national conferences. Unfortunately the most recent meeting in which we had planned to present our results was cancelled due to the pandemic. In conclusion, we managed to complete our studies before the pandemic but the dissemination has been affected by the current situation.