Depression will become the second most debilitating disorder by 2020 and the largest contributor to disease burden by 2030 (WHO, 2012), placing an ever-increasing strain on the individual, society and, importantly, the health care system. At the moment, only one quarter of the people struggling are getting the help they need (Olfson M. et al, 2016). So, how can we reduce barriers to treatment and implement interventions that are accessible, affordable, and effective?
Cognitive Behavioural Therapy (CBT) is a well-researched and clinically supported treatment option for those who are struggling with depression (NICE, 2009). Face-to-face, individual CBT treatment is highly resource-intensive, and it may prevent certain sub-groups of clients from accessing help (e.g. those with social anxiety, restricted mobility or those living in remote areas). In recent years, several meta-analyses have presented evidence that the effectiveness of group, guided self-help and telephone CBT are comparable to individual CBT for the treatment of adult depression (Karyotaki E. et al, 2017; Huntley AL. et al, 2012; Mohr D. et al, 2008). For the first time, the current network meta-analysis synthesises the available evidence, integrating direct and indirect comparisons between CBT delivery formats (Cuijpers P. et al, 2019).
The authors focussed on acute, adult depression when selecting randomised controlled trials (RCTs) from a specialised, exhaustive database (A database updated yearly from PubMed, PsycINFO, Embase, and the Cochrane Library. Literature search dates encompassed January 1, 1966, to January 1, 2018). In the selected studies, depression was established by a diagnostic interview or via a validated self-report measure, and one form of CBT delivery was compared to another form, or a control condition.
The authors used a random-effects model to pool the data of the 155 studies that were included. They compared the effectiveness (severity of depression, effect size indicated by the Standardised Mean Difference (SMD)) and acceptability (dropout from treatment, value measured by the Risk Ratio (RR)) of five forms of CBT delivery (individual, group, telephone, guided self-help, unguided self-help) and three control conditions (waiting list, care as usual or pill placebo) via pairwise and network meta-analyses.
They assessed the certainty of evidence and also tested for:
- risk of bias
- loop inconsistency
- publication bias
Finally, they conducted three sensitivity analyses, as well as a 3 to 12 month follow-up examination of the results to validate their findings.
Results of the network meta-analysis showed that individual, group, telephone, and guided self-help CBT were all similarly effective, apart from a small significant advantage of group compared to guided self-help CBT. Individual, group, telephone, and guided self-help CBT were also significantly more effective than unguided self-help CBT (SMDs ranging from 0.34 to 0.59, small-medium effect) the waiting list (SMD, 0.87 to 1.11, large effect) or care as usual (SMD, 0.47 to 0.72, medium effect). Still, unguided self-help CBT was proven to be more effective than the waiting list (SMD, 0.52, medium effect), but not more effective than care as usual (SMD 0.13). Studies comparing pill placebo did not provide significant results due to low statistical power.
The authors also examined the long-term comparative effectiveness of the CBT treatments at 3 to 12 months follow-up and found that individual, group, guided self-help and telephone CBT were all significantly more effective than care as usual or being on the waiting list.
The ranking of CBT treatment formats for acceptability (risk ratio, compared with care as usual):
- telephone (1.06)
- individual (1.03)
- group (0.99)
- unguided self-help (0.86)
- guided self-help (0.72)
The network meta-analysis found no significant difference in acceptability between individual, group and telephone CBT compared to one-another. Despite its comparable effectiveness, guided self-help produced higher dropout rates compared with all other CBT delivery options and control conditions.
Validity and certainty of evidence
Overall, the nodes of the network had sufficient numbers of studies and were well connected. The distribution of potential effect modifiers, stemming from trial and patient characteristics, was similar across the network; i.e. the transitivity assumption was upheld. There was no significant inconsistency, only low(er) power in some loops of the network; telephone CBT was compared in much fewer trials, and pill placebo was included only in two studies.
Importantly though, there was substantial risk of bias as well as considerable publication bias among the RCTs included. Furthermore, heterogeneity was moderate to high in all pairwise comparisons, and certainty of evidence (GRADE score) was moderate to low across the network for most comparisons. Nonetheless, the results of the sensitivity analyses (one of which only included low risk studies) still supported the findings of the original network meta-analysis.
- The authors concluded that group, telephone, and guided self-help CBT are effective and could be regarded viable alternatives to individual CBT for the treatment of acute, adult depression.
- They also inferred that the low acceptability of guided self-help CBT needs more examination, as its effectiveness compares to that of individual treatment.
- They hope that their results will inform future policy globally, since implementing alternative forms of CBT will make treatment more accessible to patient populations in a wider variety of settings.
Strengths and limitations
This is the first network meta-analysis on the topic, integrating a large number of studies from an extensive period (1977 – 2018).
The issue of selective outcome reporting and publication bias in studies examining the effectiveness of psychotherapies has been highlighted in Mental Elf blogs recently; researchers over-reporting effectiveness or not measuring negative outcomes (Gant E., 2019; Laws K., 2019). For this reason in particular, the authors could have solidified the validity of the present network meta-analysis by searching unpublished literature and by extending the analysis to comparing the CBT delivery formats on adverse effects (e.g. relapse, self-harm, or rates of suicide).
Whilst CBT was clearly defined, the authors found that most forms of treatment also included other skills and behavioural training alongside CBT, further contributing to the clinical as well as the methodological heterogeneity. Reflecting on this issue, care as usual or waiting list may have also varied considerably depending on the mental health setting, era of publication, or sociocultural context in which an RCT was conducted. Considering the type of mental health, medical, or social services accessible could have further nuanced the results.
On the note of nuancing results, the network meta-analysis excluded studies that included patients with anxiety or substance use disorders: both diagnoses often comorbid with depression (NICE, 2009). Admittedly, having well-defined inclusion and exclusion criteria is crucial, but it does lead to the creation of “manufactured” patient groups and raises the question: With whom in mind are we generalising and implementing the findings? Which subgroups of depressed patients are we omitting when looking for effective and acceptable CBT treatments that will then inform policy and practice, and impact the intervention they receive?
Implications for practice
The National Health Service (NHS) is catering to high demand with limited resources. The current network meta-analysis is validating a way forward that could simultaneously be a solution to financial considerations and to reducing waiting lists whilst aligning with the current Improving Access to Psychological Therapies (IAPT) stepped care model (NICE, 2009). The comparable effectiveness and acceptability of the various treatment deliveries could mean that CBT for depression will be easier to implement across a variety of clinical populations and settings.
This could prompt the uncomfortable (or inflammatory) question: Are CBT therapists redundant after all? – which is by no means implied by the paper. Instead, it prompts policy-makers and practitioners to consider equally viable alternatives to improve access while not compromising on effectiveness. Depression is costly, debilitating, and deadly and the shared aim is to make effective and acceptable therapies accessible to all those struggling – which is where this paper fits in. Nevertheless, there is reason to be cautious and to avoid implying that all forms of CBT are equally effective and acceptable for every adult diagnosed with depression. As we are gaining more understanding of depression not as an individual disorder but rather an umbrella term for different subtypes of “depressions” (Chekroud AM. et al, 2017), one avenue forward may be to outline subgroups of patients for whom one CBT format is effective and acceptable, but not another and vice versa, to produce more tailored suggestions to individual patients for CBT delivery.
“It’s the relationship that heals”– wrote the renowned psychiatrist and therapist (Yalom ID., 1989) – which may be especially true when it comes to the effectiveness and the acceptance rates of self-help therapies. On the one hand, guided self-help CBT was found to be an effective treatment but the acceptance rates were particularly low; and on the other hand, unguided self-help CBT was significantly less effective, but was not significantly less acceptable than other CBT treatments. This could imply that the lack of therapeutic alliance in unguided self-help CBT impacts on the effectiveness significantly whereas limited contact with a therapist in guided self-help CBT alters the therapeutic alliance, for it is not cultivated the same way as it would be in face-to-face sessions – still delivering effective treatment, but retaining significantly fewer patients.
To unpack these findings and improve outcomes across depressed populations, it would be important to investigate these results further.
King’s MSc in Mental Health Studies
This blog has been written by a student on the Mental Health Studies MSc at King’s College London. A full list of blogs by King’s MSc students from can be found here, and you can follow the Mental Health Studies MSc team on Twitter.
We regularly publish blogs written by individual students or groups of students studying at universities that subscribe to the National Elf Service. Contact us if you’d like to find out more about how this could work for your university.
Conflicts of interest
Cuijpers P, Noma H, Karyotaki E. et al (2019) Effectiveness and Acceptability of Cognitive Behavior Therapy Delivery Formats in Adults With Depression: A Network Meta-analysis. JAMA Psychiatry 2019 76(7) 700-707. [JAMA Psychiatry abstract]
Chekroud AM, Gueorguieva R, Krumholz HM. et al (2017) Reevaluating the Efficacy and Predictability of Antidepressant Treatments: A Symptom Clustering Approach. JAMA Psychiatry 2017 74(4) 370–378.
Gant E. Psychotherapy for adult depression: is it as good as it’s cracked up to be? The Mental Elf, 13 March 2019.
Huntley AL, Araya R, Salisbury C. (2012) Group psychological therapies for depression in the community: systematic review and meta-analysis. Br J Psychiatry 2012 200(3) 184-190.
Karyotaki E, Riper H, Twisk J. et al (2017) Efficacy of self-guided internet-based cognitive behavioral therapy in the treatment of depressive symptoms:a meta-analysis of individual participant data. JAMA Psychiatry 2017 74(4) 351-359.
Laws K. The Trial: pharmacotherapy versus psychotherapy for schizophrenia – how do trials compare? The Mental Elf, 19 Nov 2019.
Mohr D, Vella L, Hart S. et al (2008) The Effect of Telephone‐Administered Psychotherapy on Symptoms of Depression and Attrition: A Meta‐Analysis. Clin psychol (New York) 2008 15(3) 243-253.
NICE (National Institute for Health and Care Excellence) (2009) Depression in adults: recognition and management. Clinical guideline CG90.
Olfson M, Blanco C & Marcus SC. (2016) Treatment of adult depression in the United States. JAMA Internal Medicine 2016 176(10) 1482–1491.
WHO (World Health Organisation) (2012) Depression – A Global Crisis: World Mental Health Day, October 10 2012. World Federation for Mental Health 2012 32.
Yalom ID. (1989) Love’s executioner, and other tales of psychotherapy. HarperPerennial (New York)