Targets and outcomes of psychological interventions: implications for guidelines and policy

P. Cuijpers' review1 on targets and outcomes of psychotherapies for mental disorders is pertinent to the World Health Organization (WHO)'s guidance on psychological interventions. The WHO adopted in 2007 a formal methodological approach to making guidelines. Since that time, it has produced a range of mental health guidelines, including those that cover psychological interventions1, 2, 3.

As background, the WHO guidelines development process follows the Grading of Recommendations, Assessment, Development and Evaluations (GRADE) framework4. The process starts with producing a protocol for guideline development that describes a proposed independent group of experts called the Guidelines Development Group (GDG). Group membership is only confirmed after declaration of interests are reviewed. Scoping questions – for example on the effectiveness of psychological interventions – are proposed to and reviewed by the GDG.

Questions are formulated using PICO format, where P stands for population, I for intervention, C for comparator, and O for outcome. Most PICO questions list multiple outcomes. Based on the scoping questions, systematic reviews are commissioned, except when a relevant recent review already exists. The evidence is synthesized – which typically involves meta‐analysis – and then graded to communicate the certainty of the evidence, giving a transparent indication of how certain the reported effects likely are. Beyond evidence for the effectiveness of the interventions, there is systematic consideration for questions of balance of benefits versus harm, values and preferences, equity and human rights, acceptability, feasibility and resource implications. Informed by these considerations, the GDG then agrees on recommendations, which are subject to external review before finalization. The work is under the oversight of WHO's independent Guidelines Review Committee.

The above described process is not unique to WHO, and worldwide agencies and associations increasingly use similarly stringent and transparent processes of guidelines development involving answering PICO‐formulated scoping questions, though WHO guidelines are likely unique in combining a global scope with independence from industry and other external pressures.

Cuijpers' review speaks to the outcome component of PICO questions in WHO guidelines. For example, in 2013, the WHO completed a guideline on the management of conditions specifically related to stress, which included the following scoping question with four outcomes: “For adults with post‐traumatic stress disorder (P), do psychological interventions (I), when compared to treatment as usual, waiting list or no treatment (C), result in reduction of symptoms, improved functioning/quality of life, presence of disorder or adverse effects (O)?3. The GDG was asked to rank the listed outcomes according to importance using the GRADE levels (critical, important, not important). Both symptom reduction and improved functioning/quality of life were ranked as critical, while the other two outcomes were ranked as important.

In randomized controlled trials (RCTs), researchers target just one primary outcome, while, in start‐of‐the‐art guidelines processes involving PICO scoping questions, there often are multiple critical outcomes. This makes the findings of Cuijpers' review pertinent to guidelines and policy development. Given that most trials in psychological interventions target symptoms as primary outcomes, it is not surprising that the vast majority of evidence is for that outcome. But what about other outcomes? Cuijpers' review shows that in psychological intervention research there is much less data on outcomes other than symptoms. He argues convincingly that there is specifically a need for more evidence on functioning and patient‐defined outcomes, which in the context of psychological intervention research may be better referred to as person‐defined outcomes.

Over the last ten years, the WHO has made at least ten recommendations on psychological interventions through its guidelines processes. For all these recommendations there was meta‐analyzed evidence available on symptom reduction, but for none of these recommendations there was such evidence available on functioning. For functioning, the solution to this gap is straightforward: it would involve the routine adoption of functioning as an outcome in psychological intervention trials.

As crucially highlighted by Cuijpers, science will progress quickest if the same outcome measure is used across trials. Which outcome measure should that be for functioning? Ideally, a multidisciplinary group of stakeholders should propose what agreed scale should be consistently used to measure the functioning outcome across trials. I believe that they would propose the WHO Disability Assessment Schedule (WHODAS)5 for this.

The WHODAS may be identified as routine outcome measure in psychological intervention outcome research among adults, because it is the only measure of functioning that: a) has population norms and validation data across different countries; b) is well‐understood both internationally and – through its inclusion in the DSM‐5 – in the country that produces the most psychological treatment outcome data (i.e., the US); c) is already been used successfully in a range of major international studies6, 7, 8; d) provides data that can be easily analyzed for cost‐effectiveness studies6, including possible conversion into population‐level outcomes such as quality adjusted life years (QALYs), which is important for policy making; and e) is used in research across different areas of health, making improvements in its scores interpretable by an audience beyond mental health experts.

Cuijpers also emphasizes the need to collect data on the perspectives of those who are meant to be helped by the intervention, the so‐called patients, clients, service users, consumers, or people with lived experience. Though WHO guidelines take the perspectives of these and other key stakeholders into consideration, so far the WHO GDGs have not listed person‐defined outcomes as outcomes in PICO questions, likely because of the absence of a strong research tradition to collect such data.

It is hoped that this may change in the future. Indeed, at the WHO we are promoting the use of person‐defined outcomes through their routine inclusion in our own RCTs of psychological interventions among communities affected by adversity8. Again, the consistent use of the same outcome measure will be important. At the WHO we currently use the Psychological Outcome Profiles (PSYCHLOPS)9 in many of our trials, and the experiences thus far are positive.

Showing effects on a person‐defined outcome measure is helpful to convince skeptics of etic approaches10, who in some countries may include local policy makers, that a suggested psychological intervention is locally meaningful.