Evaluation of Internet-based cognitive behavioral self-help treatments for persecutory ideation and auditory verbal hallucinations
Final Report Abstract
1. Summary a. Conducted research work in relation to the objectives, milestones and hypotheses Finalization of treatment programs. The aim of the study was to evaluate an Internet-based guided self-help treatment for people with psychosis with a randomized controlled trial. The treatment was developed in accordance with the state-of-the-art knowledge regarding the underlying psychological mechanisms of the formation and maintenance of psychotic symptoms, particularly regarding auditory verbal hallucinations and persecutory delusions. Contributors from both centers (Bern and Hamburg) wrote and revised the eleven modules of the program1, and an external, independent expert with profound experience in psychotherapy for people with psychosis commented on the core modules (Dr. Eva Heibach, Germany). Additionally, we created video illustrations (e.g., of the vicious circle of sleep disorders), designed interactive worksheets (e.g., ABC protocol) and amended the self-help texts with photos. Technical infrastructure setup. The technical infrastructure was mainly set up by Stefan Westermann (Internet platform) and Pascal Zaugg (App). The development of the App — which was supposed to be compatible with iOS (Apple) and Android — took considerably more time than expected. Particularly, the implementation of reminders to use the mobile worksheets and the online synchronization of data was challenging. However, the final App was fully functional and could be maintained well during the active study phase. In addition, the questionnaire batteries and other prerequisites for the study were realized in the beginning of the project (using the Unipark platform), and a wiki for the study coordination was set up. Also, the ethics approval was obtained form the Cantonal Ethics Committee Bern (KEK 031/41). Participant recruitment and diagnostic assessments. In parallel to the technical setup, diagnostic raters and self-help guides were trained in both centers. Measure were taken to ensure agreement between raters within and between study centers, for example by providing extensive training and documentation (study wiki platform) and by ratings of a standardized, pre-recorded diagnostic interview. Guides and raters were training in the management of adverse events such as suicidality with trainings, role plays and checklists. The recruitment started on December 6, 2016; that is, about 1 year later than initially planned, due to the delays in the app development and delayed availability of personnel. About 5’000 interested persons visited the study website, from which 758 applied to take part in the study by giving informed electronic consent. Of those, 101 fulfilled all inclusion criteria (13%) and were randomized into the study conditions. In the intervention group, 43 of 50 participants had sufficient primary outcome data (86%), and in the wait list group, 46 of 51 participants (90%). Thus, the attrition rate was very low. Because the recruitment process was slower than expected, we had to prolong the active recruitment phase and applied for a cost-neutral extension of the SNSF project (which was granted). We took several measures to increase the recruitment rate, including 1) written and 2) electronic advertising in inpatient clinics, 3) flyers, 4) ads in Internet forums, 5) advertising in (offline) self-help groups and associations, 6) talks in inpatient clinics, 7) Google AdWords, 8) use of mailing lists, and 9) an article in a journal for people with mental disorders and their relatives (‘Psychosoziale Umschau’). Due to our extensive efforts, we were able to recruit 101 participants. This is less than the power analysis suggested for the detection of a medium effect size (namely, target N=128), but is still enough power (89.9%) to detect a medium to large effect (f=0.325). Guidance of participants during self-help treatment: The guides were supervised by S. Westermann in biweekly meetings. Guides from the study center in Hamburg took part via phone conference. Follow-up assessments. The follow-up assessments began in early 2017 and were completed on April 4th, 2019. Due to very extensive efforts, a high rate of follow-up assessments could be reached. We sent email reminders to overdue participants and used phone calls or mobile phone text messages to reach participants (e.g., when the emails with assessment links were moved in the SPAM folder of participants). Data analysis and main publication, as well as preparation of publications with further analyses. We presented preliminary analyses of the dataset (without the follow-up assessments) on several conferences. Due to the slow recruitment process and the resulting long recruitment window, the last follow-up assessment was not due until April 2019. In the last months, the data was checked and cleaned if necessary, and the main dataset has been compiled. Currently, we are analyzing the dataset to answer the main hypothesis. The manuscript that reports the main outcomes including the follow-ups is in preparation and will be submitted for publication during the summer 2019. In addition, three other manuscripts are in preparation (regarding the adherence, the effect of mindfulness, and the change of symptoms over time). Grant preparation. During the ongoing project, Stefan Westermann successfully applied for another grant (SNSF Advanced Postdoc.mobility fellowship; #177678). This project investigates how the adherence to Internet-based interventions can be improved. b. Main research results and explain their relevance in relation to research output Preliminary analyses (subsample of N=89) indicate that the randomization has been successful, so that there are no baseline differences between the intervention group and the wait list group regarding sociodemographic and clinical variables (all ps > 0.18). The majority of the participants where in an acute state during study intake (51%). With regard to the main outcome measure ‘auditory hallucination severity’, the results were in line with the hypothesis: both in the intention-to-treat (ITT) and per-protocol (PP) analyses the interaction effect of group and time was significant (F(1, 74.8)=8.47, p=0.005, d=0.44, and F(1, 61.4)=17.65, p<0.001, d=0.61, respectively). With regard to ‘paranoid ideation’, the interaction of group and time was significant for the PP analysis, F(1, 61.4)=5.36, p<0.024, d=0.47, but not for the ITT analysis, F(1, 75.1)=1.98, p=0.164, d=0.25. The interaction effect was not significant for the third primary outcome, the Positive and Negative Syndrome Scale, F(1, 69.6)=0.003, p=0.959, d=0.01. We reported preliminary results of the study in seven talks at conferences (see section 2c) and are preparing several manuscripts for publication (see section 2b). In addition, we successfully published related papers such as the study protocol, a narrative literature review, and the findings of the project’s pilot study (see section 2a). c. Major deviations from the research plan with justification There were three major deviations from the initially planned research plan. First, the study setup took more time than expected. Particularly, the programming of the multi-platform smartphone app was challenging. In addition, the PhD student in the project (Nina Rüegg) was only fully available starting in July 2016 and worked only part time prior to that date (as student assistant, 30%). Therefore, the start of recruitment was delayed. Second, the planned split of the study into two separate programs and thus two randomized controlled trials (RCTs) was reconsidered in the beginning of the project. Due to the significant overlap of the program contents and the target groups, the PIs decided to consolidate the two RCTs into one RCT. Therefore, the problem of competition between RCTs for participants could be avoided. In retrospect, this decision has been very valuable, as the recruitment has been slower than expected and we have one RCT with N=101 and not, for instance, two RCTs with N=50 and N=51. Third, the supervision of the guides and raters has been undertaken by Stefan Westermann and not by an external therapist (saving 18’000 CHF of the budget), and the iPods for the participants with a smartphone were not purchased, because there was no demand for this at the side of the participants (saving 4'700 CHF of the budget).
Publications
- (2017). Internet- und mobilbasierte Interventionen bei Schizophrenie. Verhaltenstherapie, 27, 181-189
Westermann, S., Moritz, S., & Berger, T.
(See online at https://doi.org/10.1159/000479335) - An online self-help intervention for people with disorders: Development, pilot testing, and participant characteristics. 48th Annual Meeting of SPR
Nina Rüegg
- (2018). An Internet-based Intervention for People with Psychosis (EviBaS): Study Protocol for a Randomized Controlled Trial. BMC Psychiatry, 18
Rüegg, N., Moritz, S., Berger, T., Lüdtke, L., & Westermann, S.
(See online at https://doi.org/10.1186/s12888-018-1644-8) - (2018). Metacognitive Training Online: A Pilot Study of an Internet-based Intervention for People with Schizophrenia. Journal for Neuropsychology, 29, 35-47
Rüegg, N., Moritz, S., & Westermann, S.
(See online at https://doi.org/10.1024/1016-264X/a000213) - Internet-basierte Selbsthilfe für Menschen mit Psychose. 36. Symposium der Fachgruppe Klinische Psychologie und Psychotherapie der DGPs, Landau
Stefan Westermann
- Internet-basierte Selbsthilfe für Menschen mit Schizophrenie: Treten unerwünschte Nebenwirkungen auf? 36. Symposium der Fachgruppe Klinische Psychologie und Psychotherapie der DGPs, Landau
Nina Rüegg
