Evid Based Mental Health 16:50 doi:10.1136/eb-2013-101297
  • Therapeutics

Adjunctive CBT increases response in pharmacotherapy-resistant depression in primary care


Question: How effective is adding cognitive behavioural therapy (CBT) to pharmacotherapy-based usual care for primary care patients with treatment resistant depression?

Patients: A total of 469 adults (average age 49.6 years, 72% female and 44% in paid employment) with treatment resistant depression (meeting ICD-10 criteria for depression, treated with adequate dose of antidepressants for ≥6 weeks but still with Beck depression inventory (BDI) score ≥14). For more than half of the participants (59%) the current depression episode had lasted 2 years or longer, and 70% had taken their antidepressant for over a year. People with bipolar disorder, psychotic disorder, or major alcohol or substance abuse problems were excluded, as were pregnant women, those unable to complete study questionnaires, those currently receiving CBT, other psychotherapy or secondary care for depression, or those who had received CBT in the past 3 years.

Setting: A total of 73 general practices in Bristol, Exeter and Glasgow, UK; 2008–2011.

Intervention: CBT plus usual care or usual care alone. CBT was provided as 12 individual sessions of 50–60 min, with up to 6 additional sessions allowed based on therapists’ judgement of clinical need. Therapists used CBT for depression treatment manuals, with expansion as needed to address treatment resistance. Therapists had at least 1 day of training specific to the trial and received weekly supervision from experienced CBT supervisors. Usual care included antidepressants, and participants were expected to continue with these during the trial. There was no restriction on usual care, and participants could be referred for counselling, CBT or for secondary care as needed.

Outcomes: Primary outcome: response at 6 months (≥50% reduction in BDI score from baseline). Analyses were adjusted for baseline measure of the outcome and stratification and minimisation variables (see notes).

Patient follow-up: 90% at 6 months and 84% at 12 months.


Design: Pragmatic randomised controlled trial (CoBalT trial).

Allocation: Concealed.

Blinding: Unblinded.

Follow-up period: 12 months (includes treatment period).

Main results

Mean duration of the CBT intervention was 6.3 months. Adding CBT to usual care significantly increased the treatment response compared with usual care alone at 6 months (response 46% with CBT plus usual care vs 22% with usual care alone; adjusted OR 3.26, 95% CI 2.10 to 5.06; p<0.001). Further adjusting these results for imbalances at baseline were reported to have a little effect (data not shown). Adding CBT to usual care also significantly increased the treatment response compared with usual care alone at 12 months (response 55% with CBT plus usual care vs 31% with usual care alone; adjusted OR 2.89, 95% CI 2.03 to 4.10; p<0.001; repeated measures analysis).


Adding CBT to the usual care in people in primary care whose depression has not responded to antidepressants is effective at increasing treatment response for up to 1 year.

Notes: Randomisation was stratified by the centre and minimised according to the baseline BDI score, previous antidepressant treatment, duration of current episode and whether the practice had a counsellor.

Abstracted from

[CrossRef][Medline][Web of Science]Google Scholar


  • Sources of funding: National Institute for Health Research Health Technology Assessment.


  1. Derek Hopko
  1. Department of Psychology, The University of Tennessee, Knoxville, USA

In past decades, through methodologically sound randomised controlled trials, cognitive behavioural therapy (CBT) has increasingly become solidified as an empirically validated treatment for depression, even among patients with severe and chronic depressive episodes. Various antidepressant medications also have demonstrated efficacy, with selective serotonin reuptake inhibitors and serotonin-norepinephrine reuptake inhibitors showing particular promise as frontline treatment strategies. Highly understudied, however, are the potential benefits of augmenting pharmacotherapy with CBT in patients with treatment-resistant depression.

Wiles and colleagues’ CoBalT study is highly innovative, in being the first large-scale randomised trial to address this issue. The clinical and social significance of this study is highlighted, in that less than 50% of individuals treated with pharmacotherapy (alone) exhibit adequate treatment response.1 In a methodologically stringent examination, the authors demonstrated that augmenting usual care (that included pharmacotherapy) with CBT significantly enhanced treatment response in individuals with treatment resistant depression. The authors should be commended on efforts to control potentially confounding variables and stratification procedures. Study findings are highly intriguing and support the incremental benefits of providing CBT to a pharmacotherapy regimen. Given that the study was conducted exclusively with a female Caucasian sample, exploring the generalisability of findings to other patient samples will be critical including potential applications to patients with more complicated psychiatric and medical comorbidities. Moreover, although study findings are provocative, a highly serious question relates to the actual feasibility of providing 12–18 (h long) sessions of CBT in traditional primary care settings. Although the authors comment that the current mental health climate in England and Australia may be conducive to such an approach, economic, pragmatic and logistical issues have yet to be empirically studied. Indeed, effective dissemination and implementation will require highly systematic and collaborative efforts. More shortened interventions may prove equally effective than the standard CBT protocol provided, which might partially resolve the aforementioned issues. As stated, this is a clinically significant and pioneering investigation.


  • Competing interests None.