Opinion Piece: Are we ready to implement Global Mental Health solutions?

by Ricardo Araya

This is the first monthly report to appear in this new section of our re-launched website. The idea for this section is to create a space where we can discuss ongoing issues of interest in global mental health in a more open and critical way. I hope readers enjoy this as much as we enjoy writing this new section. These are mostly our personal views and do not represent those of the Centre or our employing institutions.

On this first one I want to briefly take a critical look at the recent push for implementation in global mental health. In my view, there are some unreasonable expectations of many in our field of what can be achieved and how long it will take. There has been a lot of activity in the field in terms of Randomised Controlled Trials (RCTs) informing on the effectiveness of mental health treatments in Low-and-Middle Countries (LMIC) over the last 15 years. Few would disagree that overall there is improvement in terms of the evidence gathered but we need to realize that we started with very low evidence and this evidence-gap between LMIC and richer Western countries still remains extensive. Using findings from richer countries to guide on decisions of implementation might be an option but we do not know if their findings can be applied in poorer settings or require adaptations or are simply useless. The fact that several of the successful mental health trials in LMIC have used adapted versions of interventions designed in the Western world might suggest that much of the knowledge acquired in studies in richer countries could be transferred once adaptations are considered. However, it is ill-advised to assume that all these interventions will work or even less so to suggest that we have a sufficiently robust evidence-base of studies undertaken in LMIC to move on to full-scale implementation.

It was therefore surprising that important recent funding calls aimed for implementation research to look at scaling up, in view of the little evidence of the effectiveness of most interventions from LMIC. All of us working on global mental health want to have more mental health interventions successfully implemented on the ground but it is important to realize that the evidence isn’t there yet. Of course, a successful RCT is not always the answer. Often, perhaps even far too often, efficacious or even effective interventions fail the test of implementation at large scale in the real world. But the sensible way to move forward, in my opinion, requires more evidence-base as we cannot rely entirely on the evidence coming from far-away settings which are evidently different from those in most LMICs.

It is indeed disappointing to reiterate that there are no more than a handful of mental health programmes that have proven to be effective and have subsequently been scaled up in LMIC. An important question remains in this so-called evidence-based world: how much evidence do we need before we advocate for implementing at a larger scale? Nobody would dispute the need to understand better the barriers and facilitators to implementation and it is possible to argue that such research questions might be better addressed using hybrid designs addressing the effectiveness as well as issues related to future implementation. Implementation and effectiveness testing could be better aligned.  But it seems far-fetched to suggest that implementation can precede effectiveness testing. There are many examples of ineffective programmes successfully implemented leading to huge waste of resources and reminding us that (dis)investing decisions are sometimes as difficult to make as decisions on investments.

The question that follows is what can be done whilst a sufficiently robust body of knowledge is accumulated. After all there is abundance of mental disorders in LMIC and people require help or treatment for these problems. As a minimum we should advocate for a critical evaluation of what has currently been implemented with a view to decide whether it is worth it to continue investing in these programmes. Some people who have been running programmes for years might feel threatened with such an approach but resources are very limited in LMIC and we need to make sure they are utilized in the most efficient way possible. Even unsuccessful things can lead to something positive. Nonetheless, I think we do not learn enough from failures and this is a real shame. As a minimum, I think basic treatment programmes with a simple monitoring system to inform future decisions is an idea that has been in the air for many years but has never materialised in a systematic way. However, if there is a push in that direction these ‘new’ information systems need to be geared more to improving systems rather than simply collecting information to fill in statistic books.

There is at least one potentially positive aspect associated with this ‘push’ for ‘implementation’ studies in LMIC: there is an opportunity to have a look at the challenges of implementation early on in the process. There are so many examples of interventions that have proven to be effective in expensive RCTs and that subsequently fail abysmally in their implementation. For instance, there are many examples in the rapidly enlarging field of m- or e-mental health where most apps currently available have failed to be implemented in a substantial way. If anything, we will learn from the difficult process of implementation under the context of struggling economies in LMIC. Personally, I felt that we needed to accumulate more knowledge before taking this step but funders thought otherwise and we should ride the wave and get the most out of this opportunity!