Mindfulness-based cognitive therapy versus treatment as usual after non-remission with NHS Talking Therapies high-intensity psychological therapy for depression: a UK-based clinical effectiveness and cost-effectiveness randomised, controlled, superiority (2025)

Introduction

About half of people with major depressive disorder do not show remission after evidence-based treatment, with symptoms remaining above clinical thresholds, and 20–30% will develop a course in which established treatments repeatedly do not lead to sustained symptom remission.1 Recently framed under the heuristic of difficult-to-treat depression,2 such disease courses are associated with ongoing functional impairment, reduced quality of life,3 and increased morbidity.4 People with difficult-to-treat depression show intense use of health-care services, with costs accumulating to high levels as the disorder persists.5 Therefore, identification of effective interventions for people who do not show remission after previous treatment is imperative.

Research on further-line treatment has almost exclusively focused on options after pharmacological treatment,6 with trials typically finding small-to-moderate effects of additional pharmacological or psychological treatments and diminishing returns with an increasing number of previous treatment attempts.7 However, this focus is narrow, given that patients state a preference for psychological treatments.8 In England, depression care is primarily organised around a stepped care model of psychological therapies implemented in UK National Health Service (NHS) Talking Therapies services (formerly Improving Access to Psychological Therapies).9 Within these services, people whose symptoms do not respond to low-intensity treatment or who present with more complex illness are offered high-intensity treatment, which consists of evidence-based psychotherapies delivered by trained and accredited psychological therapists. In 2023–24, NHS Talking Therapies offered treatment to over 1·83 million people, with those entering treatment receiving a mean of 8·2 sessions.10 However, service outcome data show that about 50% of people with depression who complete high-intensity therapies do not reach remission. Only about 10% of these patients qualify for secondary care,11 and the remainder are not offered any further specialist care. There has been little research into treatment options when people with depression have not shown remission of symptoms during a previous psychological therapy, and evaluation of the relative benefits and harms of potential approaches has been identified as a research priority by the UK National Institute for Health and Care Excellence (NICE).12

A promising candidate to serve as further-line treatment is mindfulness-based cognitive therapy (MBCT),13 an 8-week group intervention that uses mindfulness practice as a means of helping patients become better at recognising and disengaging from habitual maladaptive patterns of thinking (referred to as the ability to decentre). This training provides people with skills that remain accessible after the end of the intervention, with evidence suggesting continuing improvements in the ability to decentre as people practise mindfulness after the end of the intervention.14 Although originally developed and recommended by guidelines for the prevention of relapse in people with a history of recurrent depression whose symptoms have decreased or subsided, there is preliminary evidence that MBCT can have beneficial effects in people with current depression whose symptoms have not responded to previous pharmacological interventions.15–18 However, more evidence is needed to justify guideline endorsement as a further-line treatment, and no research has investigated the clinical effectiveness and cost-effectiveness of MBCT after non-remission following evidence-based psychological therapy.

We aimed to investigate the clinical effectiveness and cost-effectiveness of adding MBCT, delivered via videoconference, to treatment as usual in reducing symptoms of depression for people who had not shown resolution of symptoms after NHS Talking Therapies high-intensity therapy.

Methods

Results

Between April 20, 2021, and Jan 24, 2023, 349 patients were assessed for eligibility through clinical assessment, 277 of whom met initial criteria and were invited to complete baseline assessments (figure 1). 259 patients were initially found to be eligible and randomly assigned. However, 25 patients were deemed to be randomly assigned in error after retrospective checks of self-reported inclusion criteria against NHS Talking Therapies service records indicated discrepancies (in 23 cases the last PHQ-9 score recorded for the NHS Talking Therapies high-intensity treatment of the patients did not meet the criterion of being above the threshold for caseness of PHQ-9 ≥10, in one case the number of treatment sessions was recorded as below 12, and in one case the last recorded PHQ-9 score did not meet the criterion of caseness and the number of recorded high-intensity sessions was below 12). After discussion with the trial steering committee and independent data monitoring committee, it was agreed to exclude these patients from primary analyses. However, as they met all other eligibility criteria, including a PHQ-9 score above caseness at baseline assessment, it was agreed to include these participants in a sensitivity analysis for the primary outcome only. Of the 234 eligible and correctly randomly assigned participants, 118 were assigned to MBCT plus treatment as usual and 116 to treatment as usual. 19 (16%) of 118 participants allocated to MBCT either did not start the intervention (nine [7%] participants) or dropped out of treatment before receiving the minimum dose of four sessions (ten [8%] participants). Outcome data were available for 214 (91%) of 234 participants at 10-week follow-up and 203 (87%) participants at 34-week follow-up (figure 1). The final follow-up closed on Oct 17, 2023.

166 (71%) of 234 eligible participants identified as women, 65 (28%) as men, one (<1%) as other, and two (1%) preferred not to say (table 1). The mean participant age was 42·5 years (SD 13·9). 201 (86%) of 234 participants were White. Participants had experienced a mean of 6·08 previous episodes of depression (SD 13·81) before the current episode. The mean age of onset was in young adulthood (mean 20·30 years, SD 10·47). More than a third of participants reported low or moderate levels of suicidality and about two-thirds had at least one comorbid anxiety disorder at entry into the study (table 1). In terms of previous NHS Talking Therapies high-intensity therapy, participants had received a mean of 17·4 sessions (SD 4·67, range 12–40) and shown a mean improvement of 2·71 (SD 5·20) on the PHQ-9.

Treatment fidelity as assessed on the MBCT Adherence Scale was excellent (mean 31·25, SD 1·86, on a scale ranging from 0 to 34). Ratings of therapist competence on the MBI:TAC were at or above the competent level for all groups. Rates of attendance at the eight weekly group sessions ranged from 106 (90%) of 118 participants at the first session to 81 (69%) of 118 at the seventh session, with 99 (84%) of the 118 MBCT participants receiving at least the minimal adequate dose of four group sessions. Average self-reported weekly engagement in home practice (assessed using electronic versions of the MBCT practice records) ranged from 53% to 77% for formal practices (guided meditations and movement practices) and 35–72% for informal practices (shorter practices to bring mindfulness into daily life), the mean number of formal practices was 18·07 (SD 13·29), and the mean number of informal practices was 38·57 (31·57). Engagement in practice was not significantly correlated with PHQ-9 scores at entry (p=0·52 for formal practice, p=0·45 for informal practice). Inspection of resource use over the 34-week trial period indicated that most participants were taking antidepressant medication (65 [69%] of 94 participants in the MBCT plus treatment as usual group and 71 [74%] of 95 participants in the treatment as usual group) and a sizeable number of participants engaged with psychological non-trial treatment (25 [27%] of 94 participants in the MBCT plus treatment as usual group and 38 [40%] of 95 participants in the treatment as usual group). There was little engagement with psychiatrists, community psychiatry nursing, key workers, home treatment teams, or arts-related therapies (all <6%; appendix p 50), suggesting that treatment as usual was relatively homogeneous and broadly in line with NICE recommendations, which suggest antidepressants, switching to another psychological therapy, or combined treatment.

The primary intention-to-treat analysis indicated that MBCT plus treatment as usual was superior to treatment as usual alone in reducing symptoms of depression at 34-week follow-up, with an adjusted difference on the PHQ-9 of –2·49 (95% CI –3·89 to –1·09; p=0·0006; table 2), which translates into an effect size of Cohen's d –0·41 (95% CI –0·67 to –0·15). Analyses of continuous secondary outcomes at 10-week and 34-week follow-ups showed that MBCT plus treatment as usual was superior across all measures apart from the WSAS at 10-week follow-up and the Phobia Scale (table 2).

Logistic regression analyses of dichotomous secondary outcomes indicated that significantly more patients in the MBCT plus treatment as usual group than the treatment as usual alone group reached recovery, reliable improvement, and reliable recovery based on the PHQ-9 at both 10 weeks and 34 weeks (table 3). Joint rates of recovery, reliable improvement, and reliable recovery on the PHQ-9 and GAD-7, the official NHS Talking Therapies metric, were significant at 10-week follow-up but did not reach significance for recovery and reliable recovery rates at 34-week follow-up. Absolute rates of recovery at 34 weeks based on the PHQ-9 were 27% in the MBCT plus treatment as usual group and 15% in the treatment as usual alone group (table 3).

We found no evidence of trial-related or treatment-related harms. Reliable deteriorations between assessment timepoints in depressive symptoms were rare in both groups, and reliable deteriorations in anxiety symptoms occurred in about 10% of participants in both groups (table 3). Seven serious adverse events were recorded over the duration of the trial, two in participants in the MBCT plus treatment as usual group, three in participants in the treatment as usual group, and two in participants who had not been randomly assigned. Events included hospital stays for weight loss surgery, gallbladder removal, burn, infections, heart surgery, and suspected sepsis, as well as a disclosure of suicidality before trial screening. Three events occurred in women, three in men, and one in a participant who preferred not to report their gender. All events were deemed unrelated to the trial and intervention by the independent clinical monitor (appendix p 69).

Sensitivity analyses showed that the pattern of results remained unchanged when considering the effect of NHS Talking Therapies services, when considering the effect of individual therapist or therapist seniority, when including only patients who did not reach symptom levels below caseness and had shown no reliable improvement (84 of 118 participants in the MBCT plus treatment as usual group and 79 of 116 participants in the treatment as usual alone group), and when excluding any data collected outside the initial 7 days of the assessment window as well as when accounting for non-compliance using complier average causal effect analysis. The pattern of results also remained unchanged when analyses were done on imputed data and when we did a repeated measures analysis including participants with follow-up data reported for at least one follow-up time. Mean differences on the primary outcome across the different sensitivity analyses are shown in figure 2 (appendix pp 35–48).

The cost of the MBCT intervention was estimated to be £10·12 per person per session, assuming a group size of 13. Over the 34 weeks of the study, health and social care costs were lower and utilities scores higher in the MBCT plus treatment as usual group (mean costs unadjusted difference –£316·09, adjusted difference –£245·23, 95% CI –581·92 to 91·46; p=0·15; utility scores unadjusted difference 0·064, adjusted difference 0·564, 95% CI –0·004 to 0·117; p=0·069). Non-parametric bootstrapping of mean differences in costs and QALYs at 34-week follow-up resulted in 91% of scatter points situated in the cost-effectiveness plane quadrant, indicating higher effectiveness and less cost for MBCT plus treatment as usual (figure 3). The cost-effectiveness acceptability curve for the primary health and social care perspective indicated a 95% probability for MBCT plus treatment as usual being cost-effective at a willingness to pay level of less than £5000 and had an estimated 99% chance of being cost-effective at the £20 000–30 000 threshold preferred by NICE. Secondary analyses considering the widened health and social care and productivity perspective brought similar results, although effects were numerically smaller (appendix pp 59–62). Cost-effectiveness analyses using the NHS Talking Therapies reliable recovery outcome showed a similar pattern and are detailed in the appendix (pp 53, 56–57). The pattern of results remained unchanged when analyses were repeated with missing data imputed via multiple imputation (appendix pp 58, 61).

Discussion

We found MBCT plus treatment as usual, delivered via videoconference, to be superior to treatment as usual alone in reducing depressive symptomatology in people whose symptoms had not reached remission after NHS Talking Therapies high-intensity therapy, with effects of small to moderate size maintained up to 6 months after the end of treatment. Evaluation from a health and social care perspective indicated that MBCT was economically dominant, meaning that its introduction would not only increase clinical effectiveness but also reduce health and social care costs compared with treatment as usual, even when considering the additional resources required to provide the treatment.

The current sample was characterised by complex courses with early onset, high numbers of recurrence, and considerable comorbidity. These factors, plus the fact that about 70% of those who provided data at 34 weeks had also been taking antidepressant medication, suggest that non-remission after NHS Talking Therapies high-intensity therapy was unlikely to be the first failed treatment attempt and confirms that this group shows many characteristics of difficult-to-treat depression. Previous research on further-line treatments, mostly pharmacological, has found diminishing benefit from each additional treatment introduced, with reanalyses of the STAR*D trial suggesting remission rates between 8% and 19% from first to second treatment step and 2% and 6% from second to third treatment step.7 The conservatively estimated between-group difference in sustained remission of 12% in our study is similar to these rates and has the potential to meaningfully reduce disease burden. The small to medium effect size advantage of MBCT is also similar to findings from previous trials of psychological treatment in people with chronic or treatment-resistant depression, with an estimated general effect size of 0·42.33 It is unlikely that the benefits we observed for MBCT would emerge from any psychological treatment irrespective of treatment type, given that previous trials of long-term psychodynamic therapy and interpersonal psychotherapy for treatment-resistant depression found no significant advantage of these treatments compared with treatment as usual.34,35

In addition to reducing depressive symptomatology, MBCT plus treatment as usual was superior to treatment as usual alone in reducing symptoms of generalised anxiety and increasing mental wellbeing more broadly. These findings are in line with the view that mindfulness-based interventions can target transdiagnostically relevant processes, although effects on anxiety appeared more limited than effects on depression, as reflected by the fact that rates of recovery and reliable recovery based on the combination of PHQ-9 and GAD-7 did not reach significance at 34 weeks. Post-treatment gains were maintained and skills consolidated through the follow-up, indicating that patients continued to benefit beyond the intervention, which is consistent with the intention of MBCT to provide patients with portable tools that help them stay well.

When considering the ratio of costs and effects, we found that MBCT plus treatment as usual was dominant over treatment as usual alone and that the certainty of cost-effectiveness approached ceiling levels far below the conventional threshold of £20 000 per QALY used by NICE. These findings offer an unusually strong indication of cost-effectiveness from a health and social care cost perspective, arising in the context of considerable and continuing burden associated with difficult-to-treat depression and despite only small to moderate clinical effects. As such, our findings argue for the introduction of MBCT for people whose symptoms have not shown remission in NHS Talking Therapies and highlight the potential health economic benefits of extending systematic care for this group more widely.

Confidence in our findings can be high given that rates of dropout from the trial were small and definition of remission was rigorous. It is unlikely that previous non-remission was due to misdiagnosis or failure to implement evidence-based interventions correctly given the standard procedures in place in NHS Talking Therapies services. By requiring participants to have received at least 12 sessions of high-intensity therapy, the study protected against inclusion of patients whose symptoms might not have reached remission due to insufficient dosage, and results remained unchanged in a sensitivity analysis based on patients who had not reached remission and not shown any reliable improvement in their previous therapy.

A limitation of the trial is the relatively restricted follow-up period of 6 months. Furthermore, a preponderance of female and White participants limits representativeness. We did not include gender-specific analyses as the trial was not powered for subgroup analyses. There was little information on the type and quality of psychological treatment patients received as treatment as usual.

This study adds robust evidence to existing research that has brought encouraging preliminary support for the use of MBCT in patients whose symptoms of depression have not responded to pharmacological therapies. It brings combined evidence to a level at which MBCT should be considered for guideline endorsement as a further-line treatment in the UK (appendix p 3). In conclusion, MBCT should be offered as an option to patients whose symptoms have not shown remission after evidence-based therapies for depression, and efforts are needed to make this treatment more widely available.

Declaration of interests

TB is the author of a book on mindfulness-based cognitive therapy (MBCT). TB, CS, and FAR regularly provide workshops on mindfulness-based interventions. CS is a member of a training organisation that has been commissioned by NHS England to deliver MBCT training across NHS Talking Therapies services in England and co-leads the Sussex Mindfulness Centre. TB and CS are co-investigators of a programme grant evaluating an adapted MBCT course for adolescents experiencing depression. CS is co-investigator on a grant evaluating an adapted MBCT course for NHS staff. BDD is the lead of the University of Exeter AccEPT clinic, which offers courses of MBCT. AHY serves as principal investigator for several psychopharmacological trials for treatment-resistant depression and has provided paid lectures and participated on advisory boards for Flow Neuroscience, Novartis, Roche, Janssen, Takeda, Noema Pharma, Compass, Astrazenaca, Boehringer Ingelheim, Eli Lilly, LivaNova, Lundbeck, Sunovion, Servier, Janssen, Allegan, Bionomics, Sumitomo Dainippon Pharma, Sage, and Neurocentrx. All other authors declare no competing interests.

Acknowledgments

This project is funded by the National Institute for Health and Care Research (NIHR) under its Research for Patient Benefit Programme (grant reference number NIHR200750). The views expressed are those of the authors and not necessarily those of the NIHR or the UK Department of Health and Social Care. The Exeter site acknowledges support by from the Office for Life Sciences and the NIHR Mental Health Translational Research Collaboration, hosted by the NIHR Oxford Health Biomedical Research Centre. The research was sponsored by the Research Governance team at the Sussex Partnership NHS Foundation Trust. We acknowledge the invaluable advice of the trial steering committee (David M Clark, Stephen Pilling, Anne Speckens, Judy Leibowitz, Sean Ewings, and Daniel Elton), independent data monitoring committee (Dean McMillan, Nicholas Turner, and Tim Sweeney), and patient and public involvement group (Mary Ryan, Sarah Markham, and Kate Simms). We also express huge gratitude to all participating patients and the therapists who provided the mindfulness-based cognitive therapy interventions (Emma Thom, Janet Wingrove, Kay Octigan, Bridgette O’Neill, Nicola Mouat, Florian Ruths, Kim Griffiths, and Kate Feenan) and the NHS Talking Therapies services at Sussex Partnership NHS Foundation Trust, South London and Maudsley NHS Foundation Trust, Devon Partnership NHS Trust, Sussex Community NHS Foundation Trust, Surrey and Borders Partnership NHS Foundation Trust, North East London NHS Foundation Trust, Hertfordshire Partnership University NHS Foundation Trust, Coventry & Warwickshire NHS Partnership Trust, Berkshire Healthcare NHS Foundation Trust, Nottinghamshire Healthcare NHS Foundation Trust, Lincolnshire Partnership NHS Foundation Trust, the trust research and development teams, the Clinical Research Network Kent, Surrey and Sussex that supported recruitment for the trial, and the AccEPT clinic, Mood Disorders Centre, at the University of Exeter for delivering MBCT groups as part of the Exeter site in the trial. We are grateful for the helpful input and support from the research and development teams at the sponsor (Yvette Wagner and Matthew Smith) and study sites and from the director of the Exeter Clinical Trials Unit, Siobhan Creanor, throughout the study.

Supplementary Material (1)

References