Expanding the reach of psychological care: reflections on iCBT in cardiovascular populations

This invited commentary refers to ‘Effectiveness of therapist-supported internet-based cognitive behavioural therapy interventions on depression, anxiety and quality of life among patients with cardiovascular disease: a systematic review and meta-analysis’ by S. Q. Kwek et al., https://doi.org/10.1093/eurjcn/zvaf084. In 2018, Reavell et al.1 demonstrated the efficacy of CBT for depression and anxiety in cardiovascular disease (CVD) and called for more research on internet-delivered formats. Nearly seven years later, the field has matured to the point where a meta-analysis on internet-based cognitive behavioural therapy (iCBT) in this population is feasible. This systematic review and meta-analysis provide valuable insights into therapist-supported iCBT for patients with (CVD).2 The most consistent finding was as small to moderate effect in depressive symptoms, favouring iCBT. However, the apparent effectiveness of iCBT must be interpreted considering the comparator conditions used across studies. Comparisons with inactive or treatment-as-usual controls often yield larger effect sizes than those using active comparators3 (e.g. discussion forums, as in the studies by Lundgren et al.4 and Johansson et al.5) Thus, the same numerical effect size may differ in clinical relevance depending on whether the comparator is minimal care or an active condition. This underscores the challenge of synthesizing effect sizes across heterogeneous comparator types and highlights that stratifying studies by control group design could potentially yield more accurate estimates, although the current meta-analysis may not include enough studies to permit such subgrouping. From an ethical perspective, assigning patients with clinically significant mental health needs, which are known to negatively impact cardiovascular outcomes, to control groups without access to active treatment or usual care is problematic. This is particularly important given that a substantial proportion of CVD patients continue to experience symptoms of anxiety and depression even after completing cardiac rehabilitation.6,7 To address this, some studies used attention controls or crossover designs, where the control group was offered iCBT after the initial study phase. While these approaches are ethically justifiable, they limit the ability to assess long-term between-group differences, since randomization is no longer maintained over time. This makes the lack of long-term control comparisons in the current review understandable though still a limitation. Nevertheless, the long-term follow-ups in the studies by Westas et al.,8,9 which employed a crossover design and were not included in the meta-analysis, provide important complementary data. While follow-up was only feasible for the iCBT group, the results showed sustained improvements in both depressive symptoms and generalized anxiety (GAD) 12 months post-treatment, supporting iCBT as a durable intervention. The review’s findings on anxiety are more uncertain.2 Yet, it is crucial to recognize that anxiety is a heterogeneous construct, encompassing both distress disorders (e.g. GAD) and fear disorders (e.g. cardiac-related fear). Importantly, distress disorders such as GAD are closely linked to depression, often cooccurring and sharing underlying mechanisms, while fear disorders appear less strongly associated.8,10 This has implications for both intervention design and outcome measurement. For instance, Habibović et al.11 assessed anxiety using the GAD-7 in ICD patients, capturing generalized worry but potentially missing cardiac-specific fears, such as those related to defibrillator shocks, which is common in this population.12 Instruments like the Cardiac Anxiety Questionnaire (CAQ) may be more sensitive in such cases. Supporting this, Westas et al. (2023),8 found that an iCBT programme targeting depression produced sustained reductions in GAD (distress) but had more modest impact on cardiac-related fear. Moreover, the correlation between changes in depression and GAD was stronger than with fear-related symptoms.8 These findings highlight the need to distinguish between anxiety types when selecting outcome measures and designing iCBT content for example, incorporating exposure components to better address fear-based anxiety. Given that CVD patients frequently experience a combination of depression, distress, and fear, tailored or transdiagnostic iCBT formats may be especially useful to accommodate individual symptom profiles. Schneider et al.,13 for example, used a transdiagnostic approach and observed effects not only on depression and GAD, but also on cardiac fear measured with the CAQ. The meta-analysis highlights the importance of therapist support in iCBT and discusses how nurses, due to their frequent patient contact and established therapeutic relationships, may be particularly well positioned to deliver such interventions, especially in settings where access to mental health professionals is limited. Qualitative studies support this view. Patients in Westas et al.14 and Eriksson-Liebon et al.15 emphasized the value of receiving feedback from nurses who understood the somatic context of their illness and could respond in a credible and Such contextual awareness was experienced as enhancing trust and promoting motivation. Importantly, this suggests that similar benefits may be achievable when iCBT is delivered by other professionals, such as physiotherapists, provided they receive brief training in CBT and possess relevant disease-specific knowledge. This flexibility could help scale up access to evidence-based interventions. Nonetheless, close collaboration with psychologists remains essential to ensure treatment fidelity and manage complex cases. Adherence is a well-recognized challenge in iCBT, and the meta-analysis emphasized its importance for achieving treatment effects. Some CVD patients may find it difficult to stay engaged, particularly when the treatment requires time, active effort, and emotional commitment. A temporary worsening of symptoms early in the programme has been reported, as seen in one of our studies16 and in the qualitative findings by Westas et al.14 where patients described iCBT as ‘not a walk in the park’. Helping patients anticipate this trajectory and offering early support may strengthen adherence. Feedback from someone who understands how somatic illness affects daily life, such as CVD nurses, can enhance trust and engagement,14,15 which also can promote sustained participation. Moreover, iCBT should be viewed as a complement rather than a replacement for face-to-face therapy. By offering iCBT to those likely to benefit, we can increase access to in-person CBT for patients with more complex needs or for whom iCBT is not suitable. Future research on iCBT for patients with CVD may benefit from moving beyond comparisons with inactive control groups as they raise ethical considerations and offer limited guidance for implementation in real-world settings. Instead, upcoming studies could focus on comparing active formats, or exploring variations in patient involvement, for example, giving patients the option to personalize content, or comparing such approaches to a transdiagnostic programme. In these types of programmes, elements from third-wave CBT approaches such as Acceptance Commitment Therapy can be included. Although tailoring CBT to specific CVD diagnoses has shown promise, it may be neither feasible nor necessary to develop diagnosis-specific programmes. Many patients share overlapping symptoms (e.g. fatigue and dyspnoea), treatments, and psychological concerns such as avoidance of physical activity. A broader approach might therefore offer a more scalable and practical solution for implementation in routine care. Lastly, as therapist-guided iCBT continues to expand, exploring ways in which AI-based tools could support or complement therapist feedback appears to be a promising area for future inquiry. Peter Johansson (Writing—original draft [lead]). No new data were analysed or generated in support of this research.