The mental health of the population is probably worsening and we know that psychological treatments, especially cognitive behavioural therapy, can be an effective strategy to address this problem. However, finding specialised professionals and covering the costs associated with this type of intervention is complicated. In this context, the use of new technologies to deliver effective psychological treatments on a large scale is most welcome.  

This meta-analysis examines randomised controlled studies on the efficacy of apps for treating moderate and severe depression, and concludes that apps have beneficial effects compared to various control groups. The meta-analysis uses a generally sound methodology, although some statistical choices used are poorly recommended, such as using fixed-effects models when the studies analysed are very heterogeneous, or analysing together different types of control groups (e.g. treatment as usual and waiting list) that previous research has shown may have different effects.  

The results of this meta-analysis reinforce the evidence for the positive impact of new technologies in improving the mental health of the population. In fact, previous studies also support the benefits of internet-based cognitive behavioural therapy (ICBT) for other disorders, such as obsessive-compulsive disorder (OCD). In this regard, we should mention that at Hospital Clínic - IDIBAPS we are starting the first ICBT studies for OCD in collaboration with the Karolinska Institutet in Stockholm, pioneers in this field.

Beyond the methodological limitations of a meta-analysis carried out with the inclusion of studies with very small samples, short evolutions, heterogeneous in clinical and care criteria (and curiously homogeneous in the age of the population studied, young adults), and the placebo effect of being monitored, the most relevant aspect is the focus of the research.  

The use of mobile applications to intervene in depression is part of the limited biomedical discourse that interprets mental suffering exclusively as symptoms of a disease to be treated. The use of these tools to monitor and influence people's bodies and minds is not a neutral technical intervention, but rather deepens the individualistic ideology of psychological distress and neglects the social and relational determinants that condition our emotional states.  

The translation of behaviours and bodily functions into data implies the ability to control them individually, regardless of the context, and makes guilty victims out of those subjects who suffer from painful social and family situations and are unable to feel well.  

The search for an individual technological solution based on mobile apps is a further step in the dehumanisation of the understanding of mental problems and their approach, and risks becoming the discriminatory treatment of those individuals or communities who cannot afford a relationship with a professional.

Although apps in depression have been demonstrating their feasibility, acceptability and satisfaction in successive studies and meta-analyses, there was still some doubt and contradictory aspects about their effectiveness, and if so, for what type of patients, or whether this effectiveness was only limited to apps attached to the usual treatment or also independent ones; whether the effectiveness demonstrated by some studies was due to the design and active controls or not, or whether it had more to do with technical aspects such as notifications or the duration of the treatment. In short, a lot of questions that are crucial in development and studies of the kind we do with apps. The results of this meta-analysis show almost definitively the efficacy of apps in depression, taking into account the target population, the design and methodology of the study, the psychological interventions to be adapted and the notifications, among other aspects. In that sense, they help us to have clearer ideas about these and other aspects, which is of immense guidance in the development, design and execution of studies with apps, at least as far as major depression is concerned. 

Undoubtedly, the study has employed a robust and comprehensive article review and meta-analysis methodology. First, they pre-registered the review in advance, then conducted a detailed and up-to-date review of the main scientific publication databases with up to four reviewers, complemented with standardised bias detection tools and a fairly successful analysis plan to clarify aspects of apps for depression that had not been clear so far in previous meta-analyses. Undoubtedly, the authors have previously made an exhaustive review of the evidence generated so far and have had a comprehensive reading of the aspects that required greater clarity. The selection of inclusion and exclusion criteria as well as the subgroup analyses have also been judicious ensuring that the included studies have assessed the target and clinically relevant population of patients with major depression.  

While the studies included may at first glance be few, it should be noted in the balance that this is on the basis of including patients with unipolar major depression of moderate to severe severity, which is a wise move on the part of the authors, as there is greater assurance that this is the target population and minimises the risk of selection bias. The opposite, i.e. extending the severity criteria to include mild patients, would have helped to include more studies, but would risk leaving the efficacy in depression unclear. However, it is important to bear in mind the inherent limitations of app studies, which lie in the diversity of each app in terms of content and functionality, although the authors did take some of these aspects into account in the sub-analyses by groups according to characteristics. In other words, it should always be remembered that not every app for depression is the same as every other app, especially those that have not been evaluated in clinical trials such as those included in this study.