Critical Psychiatry Network




The draft makes no mention of the changes in thinking about childhood depression over time and the current controversy that exists with regard the usefulness and potential dangers of conceptualising childhood unhappiness in the more ‘medicalised’ frame as ‘childhood depression’ (Timimi, 2004). The draft does not show a capacity for reflection and critique of the impact interpreting unhappiness as depression can have not, only on individual children and their families, but more broadly, in public health in terms of how we understand and interpret childhood unhappiness. As a result the draft has nothing to say about how and why the recent trend for inappropriate use of anti-depressants as a first line treatment (Jureidini et al, 2004a) in childhood unhappiness became established. Such a superficial and narrow engagement with the subject matter has predictably resulted in a set of recommendations that have no new, creative or imaginative ideas to offer with regard helping children’s unhappiness; instead it suggests more of the same and in the sort of didactic manner that is likely to inhibit the more creative, flexible and cross-disciplinary responses needed.

The authors repeatedly acknowledge that the science and evidence base is riddled with uncertainty and subject to shifting ideas (for example with regard definition and boundaries of the disorder). Thus in evaluating the evidence the document is full of non-specific, open to interpretation statements containing words such as, ‘about’, ‘often’, ‘probably’, ‘likely’, ‘may’, ‘around’, and ‘usually’. This reflects a conceptual difficulty the authors had to struggle with of trying to fit the evidence into a linear framework that is conceptually not up to the task. For example on p108 the draft mentions that ‘around 70% of patients treated for depression have co-morbid disorders. The search for a pure form of affective illness may be futile as co-morbidity may be an intrinsic characteristic of children with affective disorders’. Despite demonstrating the ‘woolly’ nature and inability of the label ‘childhood depression’ to explain very much in isolation, the draft then gives the label enormous explanatory power, which is reflected in the certainty with which the conclusions and recommendations are written with. Thus despite arguing that co-morbidity is intrinsic to the notion, there is nothing to suggest this is being taken into account in the conclusions and recommendations. Rather than trying to produce an artificial ‘consensus’ as the draft did, surely it would be more profitable to use the evidence to develop a more multi-layered, flexible and context-rich idea of childhood unhappiness, in which the uncertainties and current critiques are acknowledged in the conclusions and recommendations. What is really disappointing about going for a ‘consensus’, however, is that the recommendations sections did not reflect the uncertainty of the evaluating evidence sections. Thus the recommendations are presented as rigid, concrete certainties with ‘must-do’ prescriptions for services. This is disturbing, as poor theory will lead to poor practice. It also poses some fundamental questions about the fitness of the authors to consider themselves ‘experts’.

Furthermore, there is a clear bias, lacking proper scientific rigour in concluding that Fluoxetine is a safe and effective treatment for this age group, and going on to recommend that other anti-depressants are tried if the patient isn’t responding to Fluoxetine (see below). Although the draft may prove helpful in curbing some of the inappropriate prescriptions of anti-depressants, without a thorough critique of the concept and utility of the diagnosis, the gateway diagnosis to prescribing anti-depressants remains conceptualised as a ‘real objective thing’. This together with supporting the use of antidepressants in this age group (all be it with safe guards) is most likely to result in a temporary dip in the number of anti-depressants prescribed, followed by a gradual increase again as the boundaries of childhood depression are slowly redrawn to enable more prescription.

We have tried to address your doubts about the diagnosis above.
















Many of the guideline authors were involved in writing a paper for the Lancet at the same time as the Jureidini paper that addressed selective publication and its influence upon both clinical decision making and evidence based guidelines. This is the first NICE guideline to be based upon both published and unpublished evidence.























‘Linear framework’ – see above.




This is contradictory., which you criticise above, makes it clear that depression is very likely to be ‘co-morbid’ and we should have an increased vigilance for depression when children have other problems.












We are not sure how you would do this ("surely it would be more profitable to use the evidence to develop a more multi-layered, flexible and context-rich idea of childhood unhappiness, in which the uncertainties and current critiques are acknowledged in the conclusions and recommendations") FROM THE EVIDENCE BASE. This guideline has been produced following internationally agreed methods which have been sent to your organisation for comment and consultation, we believe. You may disagree with the method, but we are unable/unwilling to change the method.










The recommendations on the use of fluoxetine are cautious and consistent with the meta-analysis, which included published and unpublished studies.

















We disagree with your pessimistic view of the guideline’s likely impact on prescribing.


Critical Psychiatry Network




Regarding diagnosis - whilst the draft acknowledges changing boundaries and definitions of ‘childhood depression’ it doesn’t acknowledge and appears ignorant of the critiques of the notion of ‘depression’ in general (particularly philosophical and cross-cultural, see for example Kleinman, 1977, 1987; Kleinman and Good, 1985; Krause, 1989; Currer, 1986) and ‘childhood depression’ in particular. Through adhering to an un-critiqued linear model much of this section becomes a case of ‘building castles on sand’. If the fundamental assumptions on which the arguments rest are wrong, the conclusions will also be wrong.

The guideline raises a number of concerns and problems about the diagnosis. However, the GDG do not agree that the solution is to scrap the diagnosis and the guideline (see below).


Critical Psychiatry Network




Regarding aetiology - on page 35 the draft states, "more than 95% of major depressive episodes arise with long standing psycho-social difficulties". Given that the authors appear clear about aetiology as being the result of the complexities of psycho-social life, one has to ask what extra benefit conceptualising these difficult contexts, that the authors believe causes unhappiness in children, with an individualised, linear, context depleted, biomedical label such as ‘childhood depression’, provide?

Thank you for your comment.


Critical Psychiatry Network




Regarding treatment and management in the NHS- one problem with uncritical acceptance of a poor construct is that it leads to uncritical acceptance of research built around a poorly defined construct. Thus highlighting that studies in both the UK and USA estimate that as many as 75% of children and adolescents with a clinically identifiable mood disorder remain ‘undetected’ in the community, is both an example of circular thinking (the figures are obtained from getting people to fill out questionnaires. The figures themselves have no meaning, it is our theories that give them meaning and if we start off believing that there are illnesses to be found then we will find them), and a recipe for inappropriate pathologisation and medicalisation that is likely to eventually lead to numbers of children on anti-depressants rising again.

Thank you. Please see the comments above.


Critical Psychiatry Network




Regarding consultation with service users - 25 service users were consulted but the draft does not summarise the findings. Nor is it clear what the users’ parents/carers reported. Furthermore, there is no report on where the sample came from, average age, demographic details and so on. Having consulted only 25 users without summarising the findings from this consultation, such that it is not clear what impact, if any, their views have had on the draft, makes one wonder whether their involvement was a form of ‘tokenism’.

Thank you for your comment. There was a lack of confidence in the GDG regarding the representativeness of the sample. The GDG therefore felt it unwise to base good practice points on the report and have therefore excluded the findings on this basis.

The GDG originally commissioned this piece of work as the Group did not have representation from children or young people unlike the usual membership of GDG’s. We can assure you that this was in no means tokenistic.


Critical Psychiatry Network




Regarding black and ethnic minority children - having correctly stated that there is a problem with institutional racism in CAMHS, the authors unwittingly illustrate why; as they show astonishing ignorance of the type of debates occurring within the mental health arena on the subjugation and belittling of other communities belief systems that is so typical of institutionalised racism. Thus the authors point out that the concept of ‘depression’ may be meaningless in other cultures and conclude that this leads to ‘missed’ diagnoses. Instead of encouraging practitioners to learn about and try in engage with working with other cultures’ belief system (culturally appropriate practice) they see other cultures beliefs as an obstacle to overcome in order to discover this Western concept of ‘depression’ that lies hidden by ethnic minorities (implicitly) primitive beliefs. Once this is done their (implicitly) superior Western concept of depression can be imposed on these cultures. Thus this draft will lead to further institutional racism and should, in our opinion, be considered an institutionally racist document

Thank you for your comments. The GDG do not believe that the guideline is an institutionally racist document. It would be helpful if comments could be constructive criticism, rather than destructive and dismissive.











The guideline recommends training in cultural competence (









The GDG disagree, both with your reasoning and with your conclusions.


Critical Psychiatry Network




Regarding clinical practice - paragraph in the clinical practice recommendations, calls for ‘psycho-education’ on nature, course and treatment of ‘childhood depression’. This is of course, based on the authors’ narrow consensus view of ‘childhood depression’. The way it addresses the problem of psycho-educating ethnic minority children and families (whom the draft admits may have a belief system that views depression as ‘meaningless’) is by providing written material in their language and an interpreter. This as we have pointed out in our comments on section 3.8 (above) amounts to institutional racism. In paragraph the draft recommends specific depression services should be developed. By developing services in this manner, the label becomes an entry ticket into specific services. This could result in services becoming organised around specific diagnosis, thus mirroring the managed healthcare system in the USA. As a result practice in the USA has come to revolve around the Diagnostic Statistical Manual (DSM), which, in turn has become legal, financial and economic document that is also having a pervasive impact on our common cultural beliefs about how we should view our thoughts, actions and feelings. Thus such a recommendation has moral, ethical and public health considerations as well as purely mental health ones.

Thank you. The recommendation does not recommend ‘psycho-education’. This recommendation says that it is good practice to provide children with all the information necessary for them to understand what help and treatment they may be offered, including the hazards for treatment. The GDG believe this is the right thing to do. We also believe that people whose first language is not English may need information in their own language and/or interpreters.








The recommendation has been modified to encourage specialisation. However, the GDG believes that services should be evaluated in conjunction with stakeholders, including service users/their family and people from BME groups


Critical Psychiatry Network




Regarding screening instruments - there is no acknowledgement that screening instruments may provide nothing more than an artificial measure of a disputed construct. This relates back to the problem of conceptualising children’s unhappiness in a simplistic linear framework. Getting people to fill out questionnaires is part of a mindset of circular thinking. The figures themselves have no meaning, it is our theories that give them meaning and if we start off believing that there are illnesses to be found then we will find them, which will then be used as further justification to use screening questionnaires. There is a flawed ideology within psychiatry at present that in order to know something we have to ‘measure’ it. Uncritical acceptance of this leads to poorly thought out recommendations such as that of recommending increasing use of screening questionnaires in clinical practice, which inevitably promotes a narrow linear view of childhood unhappiness.

Thank you. Please see our responses to your other comments regarding the construct/diagnosis.





We are not sure what is meant by a ‘linear framework’? You use the word linear in a number of places relating to diagnosis (etc).






The guideline is an evidence based guideline, and therefore must look at the evidence available, including those that use measurement. The guideline also makes a number of Good Practice recommendations which act as a framework for the pathway of care within which help and treatment can be provided.


Again, we are unsure what is meant by a linear view??


Critical Psychiatry Network




Regarding clinical recommendations - the recommendations will have the effect of making ‘concrete’ the idea that the linear framework of ‘childhood depression’ for understanding childhood unhappiness is valid and undisputed. Thus recommends CAMHS should routinely screen for ‘depression’ using self-report questionnaires, recommends that training needs to be provided to increase the accuracy of diagnosing depressive disorders, and recommends that tier 3 CAMHS should be trained in using interview-based instruments to detect depression. There is no evidence that money spent in such training will improve outcomes for unhappy children. Furthermore such recommendations require uncritical acceptance of the utility of the construct of ‘childhood depression’, which is disputed (see above), and may have the effect of inhibiting the more creative and ‘common sense’ approaches that use different frameworks for giving meaning to childhood unhappiness. We are also concerned about the draft highlighting particular groups, such as, homeless, refugees, and children living in institutional settings, for interventions. This runs the risk of unhelpful medicalisation of primarily social problems.

Again: ‘linear’… we don’t understand the use of this word in this context.










These recommendations are included to improve detection of children with depression. Many, if not all, children with depression have other ‘diagnosable’ problems. Therefore, it is worth screening children with other mental health problems for depression. The B grading is derived from research that shows this approach can improve the rate of detection, not that it will improve outcomes. It is assumed that detection is necessary to offer the appropriate help and treatment. We do understand that you dispute the category of depression in children (see above).


Critical Psychiatry Network




Regarding pharmacological interventions - the cautions and limiting of circumstances in which anti-depressants may be used is welcomed. However, the uncritical acceptance of the usefulness of Fluoxetine is disappointing. The main studies on which this conclusion is based have been criticised due to methodological weakness and bias reporting. At the very least this should be acknowledged. Thus the conclusion that ‘when compared to placebo, Fluoxetine produced clinically important improvement in depressive symptoms’ (p122) is disputed (Jureidini et al, 2004a+b). Emslie et al. (1997) was the first to claim a beneficial effect for an SSRI drug in children. Of 48 patients and 48 controls, aged 7-17 years with DSM III R major depression, only 34 (71%) Fluoxetine and 26 (54%) placebo completed the trial. Significant advantage to Fluoxetine over placebo on CGI was lost when patients completing the trial were analyse. Further analysis of the results was carried out using a calculation of the slopes of the rate of change over time in the two groups, a procedure based on a large number of unsupported assumptions (including that the cumulative response over time is linear). The very high dropout rates (46% of the placebo group and 29% of the Fluoxetine group) over a short study period raises questions about the reliability of the results and conclusions. Emslie et al. (2002) in 219 patients failed to achieve a statistical difference between Fluoxetine and placebo on their prospectively defined primary outcome measures. Whilst the authors openly acknowledge this lack of significance, they nevertheless go on to claim to have demonstrated efficacy. This conclusion is based on the fact that some of their secondary outcome measures did reach significance, although several did not. The authors also give a post-hoc rationalisation of why their primary outcome measure did not return a statistically significant difference. However plausible these arguments are, the fact remains that there was no significant advantage for drug over placebo on this primary outcome measure. The TADS study (2004) was funded by a US government agency but was conducted by investigators who have received industry funding. The investigators claimed to show an advantage for Fluoxetine, especially when combined with Cognitive-Behaviour Therapy. TADS consists of two separate randomised studies: a double-blind comparison of Fluoxetine (109 subjects) vs placebo (112); and an unblinded comparison between CBT alone (111) and Fluoxetine+ CBT (107). The lack of patient blinding and placebo-control in the latter group is likely to exaggerate the benefit seen in the Fluoxetine + CBT group, who receive more face-to-face contact and knew (as did their doctors) that they were not receiving placebo. Comparing results across all four groups is therefore misleading. The authors’ claim that a CBT + placebo arm would have been ‘both too expensive and too artificial to have clinical relevance’ is unconvincing. Invalid comparisons are neither value for money nor relevant. The valid finding from TADS is the lack of a statistical advantage of Fluoxetine over placebo on the primary endpoint, the Children’s Depression Rating Scale but this was not mentioned in the abstract. Despite small numbers and the exclusion of known suicidal behaviour, TADS found a trend to more suicidal behaviour (6 attempts in the Fluoxetine groups, versus 1 in the no-Fluoxetine groups), consistent with other trials of SSRIs. Putting together that result with the lack of clinically significant advantage to drug over placebo on most measures, the benefits of Fluoxetine, like all other antidepressants, are of doubtful clinical importance for children. Furthermore their conclusion that there is ‘limited evidence’ that Citalopram, Paroxetine, Sertraline and Venlafaxine improved chances of remission and response to treatment and is recommended to be considered if the patient has not responded to Fluoxetine, flies in the face of the evidence. This together with the uncritical acceptance of the evidence on Fluoxetine poses serious questions with regard the impartiality of the authors of this draft. It is also of concern that the authors conclude that ECT may be considered for young people with either ‘life threatening’ symptoms or ‘intractable’ symptoms unresponsive to other treatments. We have little knowledge concerning the possible long-term impact of such a treatment on the developing brain and have little evidence to suggest ECT is an effective treatment in this age group. Furthermore ‘life-threatening’ or ‘intractable’ symptoms are open to varying interpretations and therefore likely to result in large regional variations dependent upon individual practitioners subjective interpretations.

Thank you for your comments. The recommendations for fluoxetine are supported by our meta-analysis, which included the more recent TADS study. There are a number of problems with the TADS study, which you highlight, including that fact that the primary outcomes were no better than placebo for fluoxetine. However, the meta-analysis does show a clear, but modest, clinical (and statistical) superiority for fluoxetine over placebo. The meta-analysis also shows that fluoxetine does have a risk of increasing suicidality (which, before TADS was included was not clear). However, the recommendations are cautious and although you rightly point out that the fluoxetine plus CBT group were unblinded,, the evidence also shows that this group (fluoxetine and CBT) did not show an increased risk of suicidality. Hence the recommendations that its best to use fluoxetine IN CONJUNCTION WITH CBT (this recommendation doesn’t depend upon the evidence being derived from a double blind study).







































































In the assessment and treatment of depression in children and young people, special attention should be paid to the issues of confidentiality, the young person’s consent (including Gillick competence), parental consent, child protection, the use of the Mental Health Act in young people and the Children Act.



This is important as other studies of fluoxetine in young people/children did not show a signal for suicidality. Nevertheless, see comments above.





The guideline recommends AGAINST the use of paroxetine and venlafaxine. However, we do VERY cautiously recommend the 3rd line use of citalopram and sertraline, both of which have the smallest signals for suicidality, according to our meta-analysis.



As you can see from the guideline, the paper we published in the Lancet in april 2004, and our responses here, we have not accepted the evidence for fluoxetine (or anything else) "uncritically".



Thank you for your comments regarding the use of ECT in light of which we have amended recommendation


Critical Psychiatry Network





1. Firstly scrap the document, it’s not necessary, adds nothing new, replicates institutionally racist views, and is likely to inhibit the sort of flexible, context-rich responses to children’s unhappiness that is needed.

2. Acknowledge the subjective nature of the construct and that the utility of using a narrow linear construction of childhood unhappiness is contested.

3. Point out that the efficacy and safety of using anti-depressants in the under-18s is not yet established. Clinical practice should reflect this.

4. The document should show greater capacity for engagement with alternative conceptualisations of childhood unhappiness, that allow for greater opportunity to develop more cross-disciplinary dialogue and research, and to free the document from the potential for legitimate accusations of institutional racism.

5. Interventions should stress more context rich approaches over individualised approaches, given that the document concludes that ‘childhood depression’ occurs in the context of psychosocial adversity.

6. As a result of the technological emphasis, the guidelines fail to acknowledge the limitations and impact of the assumptions carried in the linear biopsychosocial model it uses. If the training of professional staff attached greater emphasis to understanding and appreciating the social contexts in which childhood unhappiness arises, this would make an important contribution to improving the treatment of this group of people.

Thank you for taking the trouble to comment on this guideline.

We are sorry you feel so negative about the guideline and want it scraped. We wondered if you had seen the methodology documents sent to your organisation. If you disagree with NICE methodology for producing guidelines, can you raise it with the NICE guidelines team.

We believe that the guideline does discuss the problems of diagnosis, but making recommendations about the diagnosis is outside the scope of the guideline. Also, the GDG think that scraping the diagnosis wont help the children and young people covered by this guideline.

Our meta-analysis shows some efficacy for fluoxetine only. This is made quite clear.




Please can we refer you to the NICE methodology papers. We use internationally agreed methods.

We don’t understand what you mean by ‘context-rich approaches’. Which treatments do this and have they been evaluated? Can you supply the evidence as we did not come across this in our searches.

We disagree. This guideline emphasises the need to look at and understand the context of the child and the family. The guideline’s first key recommendation is: "When assessing a child/young person with depression, healthcare professionals should routinely consider, and record in the notes, the psychological and co-morbid factors, and the social, educational and family context for the patient and family members, including the quality of interpersonal relationships, both between the patient and other family members, and with their friends and peers."


Critical Psychiatry Network





Currer C (1986), Concepts of mental and ill being: The case of Pathan mothers in Britain. In C Currer, M Stacey (eds.) Concepts of health, illness and disease: A comparative perspective. Lemington Spa: Berg.

Emslie GJ, Heiligenstein JH, Wagner KD , Hoog SL, Ernest DE, Brown E, Nilsson M, Jacobson JG (2002), Fluoxetine for acute treatment of depression in children and adolescents: a placebo-controlled, randomized clinical trial. J Am Acad Child Adolesc Psychiatry 41: 1205-1215.

Emslie GJ, Rush AJ, Weinberg WA, Kowatch RA, Hughes CW, Carmody T, Rintelmann J (1997), A double-blind, randomized, placebo-controlled trial of fluoxetine in children and adolescents with depression. Arch. Gen. Psychiatry 54: 1031-1037.

Jureidini J, Doecke C, Mansfield PR, Haby M, Menkes DB, Tonkin A (2004a), Efficacy and safety of antidepressants for children and adolescents. BMJ 328: 879-883.

Jureidini J, Tonkin A, Mansfield PR (2004b), TADS study raises concerns
BMJ 329: 1343 – 1344.

Kleinman A (1977), Depression, somatization and the new cross-cultural psychiatry. Soc Science Med 11: 3-10.

Kleinman A (1987), Anthropology and Psychiatry. The role of culture in cross cultural research on illness. Brit J Psychiatry 151: 447-454.

Kleinman A, Good B (1985), Culture and Depression. Berkley: University of California Press.

Krause I (1989), Sinking heart: A Punjabi communication of distress. Social Science and Medicine 29: 563-575.

Treatment for Adolescents with Depression Study Team (TADS) (2004), Fluoxetine, cognitive-behavioral therapy, and their combination for adolescents with depression: treatment for adolescents with depression study (TADS) randomized controlled trial. JAMA 292: 807-20

Timimi S (2004), Rethinking childhood depression. BMJ 329, 1394-1396.

Thank you.