Common mental health problems like stress, depression and anxiety affect one in six of us in any given week. Research tell us that people with mental health problems contribute £226 billion to the UK economy each year (Mental Health Foundation, 2016). A number of facets of the modern workplace can contribute to the development of mental health problems, examined in the excellent review by Harvey et al (2017) that I reviewed earlier this year.

Common mental health problems are preventable, and prevention focused workplace based psychological interventions are key to wellbeing strategies in workplaces across the globe. These include programmes to increase resilience, develop positive mental health, and address emerging or established distress. Even though mental health problems are preventable, and often easily treated in their early stages, uptake of psychological treatments at work is low.

In recent years we’ve seen a proliferation in digital mental health tools, and a challenge on the part of researchers and innovators to adequately evidence the efficacy of these new tools. These challenges both inhibit progress, and create a risk that ineffective or even damaging tools are adopted. In the workplace, there is a rational belief that digital solutions may be the answer to increasing uptake of psychological support at work.

In this new paper, Stephany Carolan and her team have undertaken the first systematic review and meta-analysis of the effectiveness of digital psychological interventions in a workplace context. Their objective was to identify the effectiveness of occupational digital mental health interventions in enhancing employee wellbeing and increasing work effectiveness. In addition, they sought to identify intervention features associated with the highest degree of engagement and adherence; two key challenges in the implementation of any digital mental health intervention.

Methods

A systematic review was conducted following the Cochrane handbook methodology for systematic reviews (Cochrane, 2017). A robust search strategy was undertaken using standard databases and wider sources including published reports from relevant agencies (e.g. NICE, HSE) and unpublished data. The search strategy linked searches for relevant mental health terms, relevant intervention (e.g. CBT, self-help etc), online modality (e.g. internet, web-based), and workplace term (e.g. workplace, employee).

A positive deviance approach was used to identify intervention measures associated with the highest levels of adherence and engagement. A positive deviance approach is an asset based approach used to identify sustainable solutions to difficult problems by identifying “uncommon, beneficial practices” (Marsh et al, 2004). Engagement was assessed by ranking the 21 included studies in percentile order by completion and intervention study group attrition. Studies in the 70^th percentile and above were reviewed to generate hypotheses about intervention features that may facilitate high levels of engagement.

Results

Included studies

Twenty-one studies were included in the review, from 1,129 identified in the search after duplicates were removed. Nine of the studies recruited from a targeted population, including individuals with depression, stress and insomnia. The other 12 studies recruited from a general population with no set psychological inclusion criteria.

Intervention characteristics

A range of digital interventions were used in the included studies:

• Over half (12/21) were based on a cognitive approach or CBT
• 3 were based on stress and coping
• 2 on mindfulness
• 1 each on social cognitive theory, positive psychology, problem solving training and acceptance and commitment training.

The mean duration of the intervention was 7.6 weeks (range 4-13 weeks).

Seventeen (81%) used a website to deliver the intervention, with 2 studies using a computer application (1 using email and 1 using a standalone computer). A range of secondary methods were deployed in studies to engage participants (including email, text messages, calls). In just over half the studies (11/21) the participants were self-guided, with 10/21 (48%) offering guidance.

Most of the studies reported on attrition and adherence (based on the completion rates for the intervention). Most studies reported the percentage of participants that had completed part or all the intervention. The mean adherence was 45%.

Seventeen of the 21 included studies used persuasive technologies:

• Tailoring (providing information relevant to specific individuals) was used by 12 studies
• Self-monitoring (enabling people to monitor themselves) was used by 9 studies
• Tunneling (leading users through a predetermined sequence of actions or events) was used by 3 studies

Meta-analysis results

• All 21 studies were included in the meta-analysis for the psychological wellbeing measures. The workplace web-based interventions resulted in significantly reduced levels of stress, depression and distress scores in the intervention group compare to the control (p<.001) with a small effect size.
• Thirteen studies were included in the meta-analysis for the work effectiveness measure. Participants in the intervention group also showed significantly greater workplace effectiveness than control (p<.003) with a small effect size.

The results clearly show that digital mental health interventions produced a small, positive effects on psychological wellbeing and work effectiveness. The review findings suggest that digital interventions have a comparable effect to other non-digital occupational interventions in both improving psychological wellbeing and enhancing engagement in the workplace. The psychological wellbeing effect sizes observed in this review were also comparable with digital interventions delivered for depression and stress management delivered in community settings (Heber et al 2017, Cuipers et al 2009).

Subgroup analyses

The subgroup analyses conducted highlighted a range of interesting findings:

• In terms of therapeutic approach, the review suggested that as currently delivered, CBT-based occupational digital mental health tools are producing comparable, but not superior results to other approaches. This is contrary to the established literature. The reviewers suggest this may be because CBT-based digital interventions are not tailored to the workplace setting. They suggest that building theoretical frameworks of occupational stress, as discussed in a previously reviewed study by Harvey et al (2017) into CBT programmes might increase effect sizes.
• No significant difference was found in terms of effect size between guided and non-guided interventions in the review. This is also out of step with conventional wisdom in the field, which suggests that as a rule guided interventions are substantially more effective than unguided ones.
• When considering the difference in effect between targeted and universal offers, the meta-analysis showed a trend toward slightly larger effect sizes in targeted populations, potentially indicating that individuals with stress, depression or insomnia may benefit more from occupational digital interventions.

Conclusions

This review demonstrated that digital interventions in the workplace can have a positive effect on psychological wellbeing and occupational effectiveness of the workforce. It identifies factor that should be borne in mind by developers and clinicians in planning new digital interventions and reframing existing face to face interventions for digital delivery, and addressing existing digital tools to workplace environments.

The positive deviance analysis suggested that increasing engagement and adherence to digital mental health programmes is key and the reviewers hypothesise that the interventions that achieve the greatest engagement and adherence are those that:

1. Offer guidance
2. Are delivered in a shorter timeframe (6-7 weeks)
3. Utilise secondary modalities for delivery (such as email and text message), potentially also including use of mobile phone apps (only 2 studies in the review used a mobile phone app as a primary modality of delivery and both were included in the high engagement group in the positive deviance analysis)
4. Use persuasive technology (such as self-monitoring, tailoring and tunneling)

Strengths and limitations

This is a robust, systematic review in a new area of great interest. The methodology is strong, with an adequate search strategy and clear inclusion and exclusion criteria which resulted in the inclusion and analysis of 21 studies.

A key area of potential challenge comes the evidence available to review. The review sought to address occupational wellbeing in relation to digital tools and for many of the included studies these were only measured by proxy, via reporting of effect on psychological wellbeing or reduction in symptoms.

When setting out to review a broad subject, it is inevitable that the studies selected will be very different in approach and outcome measures. This inevitably led to a high degree of heterogeneity between the included studies, which was anticipated, and factored into the design of the review and meta-analysis.

In a number of areas the findings of this review run contrary to established literature. This may be because there are different factors in the workplace setting which mediate the effectiveness of interventions; certainly the reviewers suggest what some of these could be. Nevertheless, further research is indicated to ascertain whether this is the case.

Implications for practice

This is a very useful review for practitioners, developers and businesses looking to make informed decisions about where to place investment in workplace delivered digital mental health tools.

In the workplace, digital tools can for some people create an opportunity to engage with a psychological therapy at free or low cost to individuals, and a lower cost than face-to-face interventions for employers. The options for employers in this regard are multiplying and it is useful to see evidence that digital tools have a comparable effect size to other tools, and an indication of features that appear to create the highest level of engagement in a workplace setting. These hypotheses need to be further explored in robust studies.

We still see two major factors at play in this review that cloud the waters; addressing these in future will enable us to create and share better interventions and better evidence.

1. Firstly, we need to address the interventions themselves. We have reached a point of evidence now where we can be certain that digital mental health interventions work. Through this review, we can now say that some of them work in an occupational setting. We do need to get better at adapting and creating interventions for a workplace setting, ensuring that the outcomes measured and the guidance and persuasive technologies are optimal for the workplace. An innovation process (or processes) that brought together the model of work-related factors in the development of common mental health problems as proposed by Harvey et al (2017) with the factors this review suggests for effective digital interventions could be very powerful. We are starting to see innovation in this field, through the likes of the Nesta/Cabinet Office Inclusive Economy Partnerships, one of which focuses on mental health at work.
2. Secondly, we need to improve the quality of primary research so that reviews and meta-analyses can include more well designed studies. This may mean identifying promising types of workplace intervention to reduce heterogeneity by comparing multiple implementations of similar interventions. It should also include a drive by researchers and developers to more rigorously integrate robust research into the development cycle of new digital mental health tools. Projects like eMen aim to improve the uptake of digital mental health tools by connecting research, innovation and services in this way.

It is interesting to note the £2m investment announced this week by the Royal Foundation in a digital mental health start-up. As more interest and investment comes to digital mental health we must take steps to ensure innovation both uses, and helps create the best evidence.

Conflicts of interest

Chris O’Sullivan works for the Mental Health Foundation. In partnership with WellMind Media Ltd, the Foundation offers Be Mindful Online, an MBCT/MBSR based digital mental health intervention that is sold to individuals, employers and health services. Be Mindful Online was not used in any of the studies included in this review.

Links

Primary paper

Carolan S, Harris PR, Cavanagh K. (2017) Improving Employee Well-Being and Effectiveness: Systematic Review and Meta-Analysis of Web-Based Psychological Interventions Delivered in the Workplace. J Med Internet Res 2017;19(7):e271 DOI: 10.2196/jmir.7583 [Free full text access] (accessed 3/10/17

Other references

Cochrane (2017) Cochrane Handbook for Systematic Reviews of Interventions.

Cuijpers P, Marks IM, van Straten A, Cavanagh K, Gega L, Andersson G. (2009) Computer-aided psychotherapy for anxiety disorders: a meta-analytic review (PDF). Cogn Behav Ther 2009;38(2):66-82. [doi: 10.1080/16506070802694776]

Harvey SB, Modini M, Joyce S, et al (2017) Can work make you mentally ill? A systematic meta-review of work-related risk factors for common mental health problems. Occup Environ Med Published Online First: 20 January 2017. doi: 10.1136/oemed-2016-104015

Heber E, Ebert DD, Lehr D, Cuijpers P, Berking M, Nobis S, et al. (2017) The benefit of Web and computer-based interventions for stress a systematic review and meta-analysis. J Med Internet Res 2017 Feb 17;19(2):e32 [FREE Full text] [doi: 10.2196/jmir.5774]

Marsh DR, Schroeder DG, Dearden KA, Sternin J, Sternin M. (2004) The power of positive deviance. BMJ 2004 Nov 13;329(7475):1177-1179.

Mental Health Foundation (2016). Added Value: Mental health as a workplace asset. Mental Health Foundation: London. Available online at https://www.mentalhealth.org.uk/addedvalue (accessed 03/10/17)

Waller R, Gilbody S. (2009) Barriers to the uptake of computerized cognitive behavioural therapy: a systematic review of the quantitative and qualitative evidence. Psychol Med 2009 May;39(5):705-712. [doi: 10.1017/S0033291708004224] [Abstract]

van Ballegooijen W, Cuijpers P, van Straten A, Karyotaki E, Andersson G, Smit JH, et al. (2014) Adherence to Internet-based and face-to-face cognitive behavioural therapy for depression: a meta-analysis. PLoS One 2014;9(7):e100674 [FREE Full text] [doi: 10.1371/journal.pone.0100674]

Work can make you mentally ill, but we still have a lot to learn about the links between employment and mental health

Digital interventions for stress: do they live up to their alleged potential?

Photo credits

• Photo by Climate KIC on Unsplash
• Photo by Vladimir Kudinov on Unsplash
• Photo by rawpixel.com on Unsplash
• Photo by Bethany Legg on Unsplash
• Photo by Jakub Kapusnak on Unsplash
• Photo by Scott Webb on Unsplash