Sources of normativity in childhood depression

Abstract

Measuring depression outcomes means making normative judgements, that is, deciding what is good or better than before, and what is bad or what may have gotten worse. It doesn’t matter if one uses a scale to measure depression or if one is simply using their clinical impression to guide them. But how do we know what is good and what is bad and whether any change is for the better? Also, how do we know what to ask, whom to ask and how to ask about such outcomes? In their important study, Krause et al. [1] did something rather unusual for our field: they asked participants of a depression psychotherapy trial about what they think had changed due to the intervention; they also asked the participant’s parents and their therapists. The psychotherapy trial was the heroic IMPACT [2] and the data in the Krause paper come from IMPACT-ME, the qualitative study nested in the bigger one. Famously, IMPACT showed no effectiveness differences between three arms: a Brief Psychosocial Intervention, Cognitive Behavioural Therapy, and ShortTerm Psychoanalytic Psychotherapy. Krause et al.’s approach is unusual because in depression trials we, the researchers, the ones who design the measurement scales and trials, are usually our own source of normativity. A typical depression scale is designed by asking experts what they believe are the items (questions) that capture the most important features of adolescent depression. A big pool of questions is created, out of which, through statistical item-reduction techniques, a final instrument (i.e. questionnaire or interview) is generated [3]. This is then handed to the patients themselves or to clinicians who interview patients and that’s how depression is measured. Alternatively, clinicians assessing the patients are asked to form a “global impression” of the adolescent and provide a rating about whether they have changed. Krause et al. show that over 90% of trials only collect information in this way and ask primarily about depressive symptoms—a sort of DSM symptom count. But are the things that we have been asking about the things that matter to adolescents, their parents or their therapists? As Chevance et al. have shown previously in adults [4], Krause also demonstrates for this age group: only to an extent. Whilst all three—adolescents, parents and therapists—often care about mood symptoms, they also care about a host of other things, such as family relationships, skills, everyday functioning, that are rarely considered in the measures that define the outcomes of depression trials. Krause et al.’s study questions our conceptions of normativity. Consider the following scenarios: “I do not feel that low any more, I got better but I still do not function properly, feel like I cannot concentrate”; or the converse, “at last I get on better with my family, but my mood has improved only a little”. Who has improved more? But many will ask, why should we care? Surely, if we only had a marker of depression as, say oncology has for tumours, all this discussion might prove to be simple academic indulgence. This view is based on at least two misconceptions. For one, such an argument on biomarkers is based on the problematic notion that science or technology are in and of themselves a source for normative judgements. It should be evident that science and its tools do not carry or generate themselves normative values, this comes from the people and the societies using the tools. It has even been cogently argued that scientific knowledge itself can only be produced because it is based on existing personal and societal constructs of normativity, of what is right vs wrong, correct vs incorrect, useful vs non-useful [5]. It is the job of good science to be transparent about such sources of normativity and their implications for practice. Also, such dreams that biomarkers will deliver us from the responsibility of addressing these questions ignores what happens in the rest of medicine. My father’s treating haematologists were tracking the mRNA expression of his NPM1 gene mutation when they were treating his leukaemia; but * Argyris Stringaris [email protected]