In depression one size does not fit all — antidepressants are not all equal
Depression is an illness and as such is amenable to pharmacological or psychological intervention. If such interventions are accurately directed by guidelines, it is commonly supposed outcomes should be favourable.
However, the National Institute of Health and Clinical Excellence states that depression is a broad and heterogeneous condition, the prognosis of which is significantly influenced by diverse biological, psychological and social factors not captured well by current diagnostic systems.1
Compare, for example, a depressed, cannabis-using, parasuicidal, 18-year-old female with a 50-year-old, affluent male with depression following a myocardial infarction. Each of them could readily meet DSM-IV (Diagnostic and Statistical Manual of Mental Disorders, 4th edition) diagnostic criteria for major depression2 but approaches to management have little in common.
Anecdotes cannot be a basis for health care policy but they do illustrate the ecological inference fallacy: the assumption that because something applies to a population then it applies to each member of that population individually.
Much current thinking about the treatment of depression can be construed as falling into such a “one size fits all” ecological fallacy: because certain pharmacological and psychological interventions have general efficacy, they should be applied routinely to all those diagnosed as depressed.
Same or different?
The assumption that antidepressants are helpful has been challenged.3,4 It has been reported that drug/placebo differences in efficacy increase with baseline severity of illness reaching clinical significance only in those at the upper end of the most severely depressed category.4
Such data have provoked critical reconsideration of the benefits of antidepressants, so let us now describe that benefit. The outcome of a typical antidepressant study is illustrated in Figure 1.
|Figure 1: A typical acute study of an antidepressant versus placebo in major depression |
Mean baseline HAMD (Hamilton Depression rating) score5 in a major depression study is typically 20–25 falling by about 40 per cent over six to eight weeks of active treatment, surpassing placebo by two, three or four points. However a change in HAMD score of three points is clinically imperceptible in an individual patient.
Data can also be looked at in terms of outcomes such as remission. In a typical study 40 per cent of patients might achieve remission on an antidepressant compared with only 20 per cent on placebo.6
Thus, it is indisputable that although antidepressants are not universally effective, neither are they mere placebos. What is debatable is not whether antidepressants are efficacious but the extent to which they are so.
Using multi-treatment meta-analysis, Cipriani et al compared 12 “new-generation” antidepressants.7 Distinct differences in efficacy, as measured by responder rate, were identified. Least efficacious was reboxetine (responder rate 48 per cent). Most efficacious were mirtazapine, escitalopram, venlafaxine and sertraline (responder rates 59–62 per cent) while fluoxetine had a responder rate of 55 per cent.
Further evidence from a Cochrane meta-analysis8 shows that compared with citalopram, the odds ratio for response was 1.49 in favour of escitalopram. This approximates to a 61 per cent responder rate with escitalopram compared with 54 per cent with citalopram.
Older analyses indicate that venlafaxine is superior to fluoxetine and clomipramine superior to paroxetine and citalopram.9–12 So, if antidepressants are not “all the same”, what are the implications?
New directions in antidepressant choice
The central problem is the nature of depression itself: there is a need for radical reflection on what we believe “depression” to be. Notwithstanding the heterogeneity of “real world” depression, subdivision into mild, moderate, severe and complex may be practically useful.1
One might argue that mild depression should not be treated with antidepressants because the pounds per quality adjusted life year (£/QALY) is too high; but severe depression should be treated with antidepressants because there they are cost-effective.
For example, benefit from escitalopram seems to correlate with baseline severity of illness.13 The more severe the depression, the more useful escitalopram. This simple logic however, breaks down under scrutiny: patients with severe depression must have gone through a trajectory from mild to moderate and, ultimately, severe depression. Thus, among those with mild depression will be some who are on such a worsening trajectory.
Should not mild cases be treated pre-emptively to avoid this? That would seem desirable but predicting such progression is difficult. This is not addressed in the pseudo-certainty of guidelines.
Fluoxetine and citalopram, as cheaper generics with no known efficacy differences from similar more expensive antidepressants, have been considered “first line” treatments in recent years. However, might it not be time to reconsider?
Many arguments could be adduced on either side: If escitalopram is more efficacious than citalopram and sertraline more than fluoxetine, should not the former supplant the latter in each case? Non-generics such as escitalopram might have higher acquisition costs but superior overall cost efficiency.
A counter argument might be that since up to 42 per cent of patients in primary care may discontinue treatment within four weeks,14 choice is immaterial — no medicine can be effective unless it is taken. Each of these arguments is based on “one size fits all” but perhaps individual patient characteristics are of central importance.
Compare and contrast the two cases referred to in the introduction: the wayward teenager and the depressed male. The clinician might wish to play for the best odds of rapid improvement by using a more expensive antidepressant with an efficacy advantage for the latter while deciding to spend less on the former, as a potentially non-adherent patient.
Putting it together
What, then, are we proposing? Are we leading to a conclusion about which is the “best” antidepressant or how one should engineer a “better” guideline? No, because to do so would be to fall back into the “one size fits all” fallacy. We do however make the following proposals:
- The dominance of citalopram and fluoxetine needs to be critically reconsidered. The direction of current evidence is that antidepressants are not “all the same”. The challenge is to know what value to ascribe to those differences: what, for example, is a 5 per cent efficacy advantage of one worth over another? A way to approach this would be in pecuniary terms: the £/QUALY and cost-effectiveness ascribed to antidepressants by NICE1 are summarised in Tables 1 and 2 (PDF 170K). Payers might reach the conclusion that citalopram and fluoxetine should remain dominant but reconsideration is important, at the very least to test current practice against the most recent evidence.
- There should be a move away from a simplistic concept of depression. Patients vary greatly in characteristics and histories, and the use of antidepressants is not comparable to the use of statins. National guidelines1,15 recognise the heterogeneity of depression but this needs to be amplified. If the question is “which antidepressant should be used first-line?” then a response might be “what type of patient are we talking about?”. This requires much thought and recognition that there is probably no one “best” antidepressant.
- The management of depression is more an art than a science: it is not simply a case of having technically accurate guidelines which are then assiduously followed by proficient clinicians. An individual case of depression may involve numerous interacting variables, influencing if, when and how to treat, and what to do in the face of insufficient improvement. Thus, until such time as there are major advances in the understanding of depression it should accepted that the judgement of the expert clinician is of central importance.
The management of depression is complicated and antidepressant use not straightforward. The body of evidence now indicates that antidepressants are not all equal. Considerable effort is required to translate that evidence into rational and cost-effective prescribing.
Chris Hawley is a consultant psychiatrist at Hertfordshire Partnership Foundation Trust and the University of Hertfordshire
Celia Feetam is a specialist psychiatric pharmacist in the department of psychiatric pharmacy at Aston University, Birmingham
1. National Institute for Health and CLinical Excellence. Depression in adults. Clinical guideline 90. Available at: www.nice.org.uk (accessed 7 November 2009).
2. DSM-IV Diagnostic Criteria. Washington DC: The American Medical Association, 1999.
3. Moncrief J. Efficacy of antidepressants in adults. Education and debate. BMJ 2005;331:157.
4. Kirsch I, Deacon BJ, Huedo-Medina TB, Scoboria A, Moore TJ, Johnson BT. Initial severity and antidepressant benefits: a meta-analysis of data submitted to the Food and Drug Administration. PLoS Medicine, 2008;5(2).
5. Hamilton M. A rating scale for depression. Journal of Neurology, Neurosurgery and Psychiatry 1960;23:56–62.
6. Thase ME, Entsuah AR, Rudolph RL. Remission rates during treatment with venlafaxine or selective serotonin reuptake inhibitors. British Journal of Psychiatry 2001;178:234–41.
7. Cipriani A, Furukawa TA, Salanti G, Geddes JR, Jiggins JPT, Churchill R et al. Comparative efficacy and acceptability of 12 new-generation antidepressants:a multitreatment meta-analysis. The Lancet 2009;373:746–58.
8. Cipriani A, Santilli C, Furukawa TA, Signoretti A, Nakagawa A, McGuire H. et al Escitalopram versus other antidepressive agents for depression. Cochrane Database of systemic Reviews 2009 Issue 2.
9. Thase ME, Entsuah AR, Rudolph RL. Remission rates during treatment with venlafaxine or selective serotonin reuptake inhibitors. British Journal of Psychiatry 2001;178:234–41.
10. Smith D, Dempster C, Glanville J, Freemantle N. Anderson I. Efficacy and tolerability of venlafaxine compared with selective serotonin reuptake inhibitors and other antidepressants: a meta-analysis. British Journal of Psychiatry 2002;180:396-404.
11. Danish University Antidepressant Group Paroxetine: a selective serotonin reuptake inhibitor showing better tolerance, but weaker antidepressant effect than clomipramine in a controlled multicenter study. Journal of Affective Disorders, 1990 18, 289-299
12. Danish University Antidepressant Group. Citalopram: clinical effect profile in comparison with clomipramine. A controlled multicenter study. Journal of Psychopharmacology 1986;90:131–8.
13. Montgomery SA, Kasper S. Severe depression and antidepressants: focus on a pooled analysis of placebo-controlled studies on agomelatine. International Clinical Psychopharmacology 2007;22:283–91.
14. Feetam CL. Medicine taking behaviour in depression, part 1. Progress in Neurology and Psychiatry 2009;13:21–24.
15. Anderson IM, Nutt DJ, Deakin JFW. Evidence-based guidelines for treating depressive disorders with antidepressants. A revision of the 2000 British Association for Psychopharmacology Guidelines. Journal of Psychopharmacology 2000;14:3–20.
Citation: The Pharmaceutical Journal URI: 10996365
Recommended from Pharmaceutical Press