Please enter the name of your registered stakeholder organisation below. If you are a non-registered organisation we strongly advise you to register via the NICE website .

If you are commenting as an individual then we recommend that you contact the registered stakeholder organisation that most closely represents your interests and pass your comments to them.

Stakeholder Organisation:

Critical Psychiatry Network

Name of commentator:

Joanna Moncrieff

Order number

Indicate number or ‘general’ if your comment relates to the whole document

Example

Full

3.4.6

45

My comments are as follows ……

Proformas that are not correctly submitted as detailed in the line above may be returned to you

1

Full and NICE

general

general

The recommendations do not reflect the considerations about the validity of the disorder or the methodological deficiencies of randomised controlled drug trials that are discussed in the full guideline. The recommendations give the impression that there is complete consensus that "depression" is a meaningful term, as now applied, whereas this is not the case. They also give the impression that evidence from drug trials is entirely reliable, in contrast to the discussion in the full guideline.

2

Full

2.1.1, 2.1.3

12-13, 18

The guideline refers to depression as a "heterogenous group of related disorders" and acknowledges that the term is used to refer to "a wide range of mental health problems". The guideline notes that "distinguishing mood changes between clinically significant degrees of depression (e.g. major depression) and those occurring normally remains problematic" (P 12). The fact that different people with "depression" have widely ranging and sometimes contrasting symptoms is also described. The Guideline commendably emphasises the "uncertainty inherent in our current understanding of depression and its classification" (P18). In view of these considerations, NICE should have considered some of the literature that suggests that the term "depression" as currently used in psychiatry is not a meaningful or valid term, and one that has been fashioned largely to market antidepressants (Dowrick C, Beyond Depression, OUP, 2004; Pilgrim D & Bentall RP. The medicalisation of misery: a critical realist analysis of the concept of depression. Journal of Mental Health, 8, 261-274, 1999; Moncrieff J. The creation of the concept of an antidepressant: an historical analysis. Social Science and Medicine, 66, 2346-2355).

3

Full

2.1.3

18

The guideline describes a diagnosis of depression as a "starting point in considering the most appropriate way of helping that individual in their particular circumstances". If NICE considers this to be the best approach, (that is helping the individual in their particular circumstances), it is not clear what having a "diagnosis" adds. Diagnosis is a group label, which implies that a certain sort of treatment will be specifically helpful. This contradicts the idea of helping individuals with their own unique set of problems.

4

Full

2.1.3

17

It is not clear why the concept of "subthreshold" depression is being introduced. Although the guideline development group does not intend it to increase antidepressant prescribing, it may have just this effect, by allowing more people to be diagnosed with some sort of depressive disorder. Where there is pressure for medical intervention, this is likely to be antidepressants, whatever the guideline recommendations.

5

Full

2.1.3

18

The guideline notes that categorising depression is inherently uncertain, and it warns against making a judgement of severity based on symptom counts, although it does not make clear what should be used instead. However, it then presents a classification based on subjective ratings of severity, and many of the most important recommendations in the Guideline classify depression according to severity.

6

Full

2.4.1, 2.4.3

22-23, 24

Therefore it is difficult to see how the Guideline can conclude that the suggested levels of non detection is such a "very undesirable state of affairs" (P23).

7

Full

7.1

204

The Guidelines state that "Systematic reviews using meta-analysis suggest that antidepressant drugs when considered individually or as a class, are more effective than placebo in the treatment of major depression". This is a weaker statement than the previous version of the guidelines, but it still does not reflect the weakness of the evidence. Two meta-analyses are cited, but others that find a clinically insignificant difference between antidepressants and placebo are not (Kirsch et al, 2002; Storosum et al, 2004). The meta-analysis of SSRI placebo controlled trials presented in the Guideline also does not support this statement.

8

Full

7.1

205

No reference is made to the problems of discontinuation studies when discussing the issue of how long antidepressant treatment should be given for. The fact that discontinuation studies are likely to be unblinded by withdrawal effects, which have not been taken into consideration, means their validity is questionable and therefore the recommendation to continue antidepressant treatment for 6 months after response is unfounded.

9

Full

7.1, 7.3

203, 207

The guideline authors make much in these sections of the assumption that people with milder depression show smaller drug placebo differences than people with more severe depression. Although some meta-analyses have shown this effect (most recently the paper cited by Kirsch et al, 2008), the evidence is conflicting and it is premature to assume that this has been established. Some evidence suggests the response may be worse in inpatients than outpatients (Moncrieff et al, 2001) and naturalistic studies of inpatients also show poor response rates (Tuma et al, 2000).

10

Full

7.5.1

209

It seems contradictory that the Guideline says that there is little data to help clarify whether antidepressants are more useful or beneficial at different levels of severity of depression. Yet the recommendations, and other statements throughout the length of the Full Guideline (see previous point) make reference to the effects of antidepressants on categories of severity. The Quick Reference Guide refers to categories of severity, without any hint that the interpretation of these is difficult or problematic.

11

Full

7.7.2

214-15

In our submission on the first version of these guidelines published in 2004, we made the point that the statement that there was "strong evidence suggesting that there is a clinically significant difference favouring SSRIs over placebo on increasing the likelihood of patients achieving a 50% reduction in depression symptoms as measured by the HRSD" contradicted the finding that the difference in change scores was so small as to be considered clinically insignificant. Since the response rates are derived by categorising the same continuous data that is used to analyse change scores, it cannot make sense for one form of analysis to produce a clinically significant difference, and another form of analysis of the same data to produce a clinically insignificant difference. Since the data is collected in continuous form, that is as scores, the analysis of symptom change scores should take precedence. We were puzzled that this point, which hsd such a fundamental bearing on the analysis, conclusions and recommendations of the guidelines was not addressed in the response to our last submission. The fact that the difference between SSRIs and placebos in terms of symptom reduction was not clinically significant should have lead to conclusions that SSRI antidepressants were not superior to placebo, and they should not therefore be recommended for any type of depression. This conclusion would be consistent with several other recent meta-analyses, which find small and clinically doubtful differences between antidepressants and placebo (Kirsch et al, 2008; Khan et al, 2002; Storosum et al, 2001).

12

Full

7.7.2

214-15

The terminology used to describe categories of depression is inconsistent and confusing. The meta-analysis employs the APAs definitions of severity- which are less stringent than most classifications. Therefore, people who would be classified as mild depression under other systems are labelled as "moderate depression" and the group in the mid range is referred to as having "severe depression." Nowhere in the document is it clearly specified how these categories relate to the categories of "mild" moderate" and "severe" depression that are used in the recommendations in the Quick Reference Guide.

13

Full

7.7.2

214-15

The analysis of continuous data finds that the greatest difference between drug and placebo occurs in the group in the mid range of severity. This contradicts the assumption made throughout the Guidelines that more severe depression responds better to antidepressants relative to placebo. Since the APA categories set severity qualifications low, this means that people normally classed as having severe depression ("very severe depression" according to APA) do not show a clinically significant difference from placebo.

14

Full

7.7.2

215

It is not clear why the analysis of symptom change scores presents results in terms of SMDs when it specifies that only HRSD scores were considered. If all studies used HRSD scores, then surely WMD should have been used.

15

Full

7.7.2

215

It would be useful to know what was the size of the difference in HRSD scores between drug and placebo groups in the analyses of depression symptom change scores that were conducted

16

Full

9.11.2- 9.12

318-21

The discussion of the potential of antidepressants to induce suicidal symptoms is important. The Guidelines identify some conflicting evidence of antidepressants increasing suicidal risk in adults, although the evidence in children and young people seems strong and consistent. There is no discussion of why there may be a different effect in children and young people versus adults and no consideration of the fact that pharmacologically this would be an unusual situation. The recommendations for increased monitoring and prescription of lower toxicity drugs seem too weak. There is plenty of time within a week for someone to become intensely suicidal and as the guidelines point out, even the supposedly safe SSRIs are associated with 6% of deaths by suicide. In any case, given the negative findings on efficacy, any possibility of a link between suicide and antidepressants should surely prompt the Guideline to discourage the use of antidepressants.

17

Full

7.3

206-8

There is an implicit suggestion in this section that positive and negative biases cancel each other out; that is the tendency to recruit milder cases counteracts the bias associated with selective publication and sponsorship. But this has not been shown to be the case and most evidence suggests that biases operate mainly in the direction of overestimating treatment effects.

18

19

20

21

22

23

24

25