A Direct-to-Public Peer Support Program (Big White Wall) Versus Web-Based Information to Aid the Self-management of Depression and Anxiety: Results and Challenges of an Automated Randomized Controlled Trial

Abstract

Background Effective help for depression and anxiety reaches a small proportion of people who might benefit from it. The scale of the problem suggests the need for effective, safe web-based public health services delivered directly to the public. One model, the Big White Wall (BWW), offers peer support at low cost. As these interventions are delivered digitally, we tested whether a randomized controlled trial (RCT) intervention could also be fully delivered and evaluated digitally. Objective This study aims to determine the reach, feasibility, acceptability, baseline costs, and outcomes of a public health campaign for an automated RCT of the BWW, providing digital peer support and information, compared with a standard website used by the National Health Service Moodzone (MZ), to people with probable mild-to-moderate depression and anxiety disorder. The primary outcome was the change in self-rated well-being at 6 weeks, measured using the Warwick-Edinburgh Mental Well-Being Scale. Methods An 18-month campaign was conducted across Nottinghamshire, the United Kingdom (target population 914,000) to advertise the trial directly to the public through general marketing, web-based and social media sources, health services, other public services, and third-sector groups. The population reach of this campaign was examined by the number of people accessing the study website and self-registering to the study. A pragmatic, parallel-group, single-blind RCT was then conducted using a fully automated trial website in which eligible participants were randomized to receive either 6 months of access to BWW or signposted to MZ. Those eligible for participation were aged >16 years with probable mild-to-moderate depression or anxiety disorders. Results Of 6483 visitors to the study website, 1510 (23.29%) were eligible. Overall, 790 of 1510 (52.32%) visitors participated. Of 790 visitors, 397 (50.3%) were randomized to BWW and 393 (49.7%) to MZ. Their mean age was 38 (SD 13.8) years, 81.0% (640/790) were female, 93.4% (738/790) were White, and 47.4% (271/572) had no contact with health services in the previous 3 months. We estimated 3-month productivity losses of £1001.01 (95% CI 868.75-1133.27; US $1380.79; 95% CI 1198.35-1563.23) per person for those employed. Only 16.6% (131/790) participants completed the primary outcome assessment. There were no differences in the primary or secondary outcomes between the 2 groups. Conclusions Most participants reached and those eligible for this trial of digital interventions were White women not in recent contact with health services and whose productivity losses represent a significant annual societal burden. A fully automated RCT recruiting directly from the public failed to recruit and retain sufficient participants to test the clinical effectiveness of this digital intervention, primarily because it did not personally engage participants and explain how these u...

TreadWill: Development and pragmatic randomized controlled trial of an unguided, computerized cognitive behavioral therapy intervention in a lower middle-income country

“…We used only one questionnaire each for assessing depression and anxiety symptom severity. This decision was made keeping in mind that filling long questionnaires online is not a pleasant experience for the users and might increase dropout (Morriss et al, 2021). Also, while including multiple questionnaires for assessing the same disorder might improve validity, it also increases the risk of getting false-positive results by chance.…”

“…In a fully remote trial of an app-based intervention, Arean et al (Arean et al, 2016) reported that 57.9% of the participants did not even download their assigned apps. Similarly, in another study involving no human contact, Morriss et al (Morriss et al, 2021) reported that only 57.3% of participants randomized to the experimental group signed up for the intervention and only 42.5% accessed it more than once. Morriss et al (Morriss et al, 2021) further reported an attrition rate of 84.9% at 3-week follow-up with the attrition rate increasing at later follow-up points.…”

“…Similarly, in another study involving no human contact, Morriss et al (Morriss et al, 2021) reported that only 57.3% of participants randomized to the experimental group signed up for the intervention and only 42.5% accessed it more than once. Morriss et al (Morriss et al, 2021) further reported an attrition rate of 84.9% at 3-week follow-up with the attrition rate increasing at later follow-up points. In another recent study, Oehler et al (Oehler et al, 2021) observed that the minimal dose was received by only 2.10% of the participants for the unguided version of iFightDepression.…”

“…Note that the unguided studies included in the metaanalyses often involved initial contact with humans for diagnostic interviews (Berger et al, 2011;de Graaf et al, 2009;Farrer et al, 2011;Meyer et al, 2015;Mira et al, 2017;Spek et al, 2007), weekly telephone contact support (Christensen et al, 2004;Phillips et al, 2014), or treatment-as-usual (Gilbody et al, 2015;Klein et al, 2016). Even minimal human contact can increase adherence to the interventions compared with a study without any such contact (Moritz et al, 2012;Morriss et al, 2021). Indeed, Fleming et al (Fleming et al, 2018) found that adherence rates observed in trial settings fail to translate into the real world.…”

Digital Mental Health Amid COVID-19

“…There were mostly contextual issues noted in recent studies. An automated randomized controlled trial of an online peer-support program (Big White Wall) versus web-based information for self-management of anxiety and depression suggested a personal approach for participant engagement to explain the benefits of this novel intervention [33]. It was proposed that user engagement and satisfaction with a digital platform (Thought Spot) be related to content relevance, ease of learning, and available features [34].…”