Selective serotonin reuptake inhibitors

The increasing use of antidepressants—particularly the selective serotonin reuptake inhibitors (SSRIs)—might be grounds for optimism because it might indicate that one of the great health problems of our age is increasingly being recognised and treated. Instead SSRIs have become embroiled in controversy over both their effectiveness and safety. As well as the doubts about efficacy, the media have fuelled concern that SSRIs may cause serious adverse effects, ranging from worsening depression to suicide. The scientific evidence shows that the media has blown the risk of suicide out of proportion.

The increase in prescribing of SSRIs has coincided with a fall in the suicide rate in many countries, implying that SSRIs are not a major cause of suicide.¹ Case-control studies—which cannot completely rule out confounding by indication—probably exclude a substantial increase in both relative and absolute risk of suicide.² A meta-analysis of individual patient data from the randomised trials is clearly necessary but has not been done. Short term randomised trials of SSRIs in children and adolescents show a modest increase in some suicidal thoughts and behaviours, but it is unwarranted to assume that this translates into an increase in the risk of suicide itself, rather than reflecting the transient increase in agitation that is a recognised adverse effect of SSRIs.³

Despite the incomplete evidence, the concern over potential adverse effects has led to regulatory authorities issuing warnings about specific groups of patients when evidence for benefit is lacking. In the US the Food and Drug Administration has asked for a statement to be included in the relabelling of several antidepressants recommending that adults and children taking these drugs should be closely observed for worsening depression and the emergence of suicidal thoughts.⁴ In the UK the Medicines and Healthcare Products Regulatory Agency has recently advised that SSRIs other than fluoxetine should not be used in children and that the dose of paroxetine should be limited to 20 mg in adults (www.mhra.gov.uk). The forthcoming National Institute for Clinical Excellence (NICE) guidelines on managing depression have been delayed until the Committee on Safety of Medicines has reconsidered the evidence on the efficacy and safety of SSRIs (www.nice.org.uk/page.aspx?o=20093). The current draft of the NICE guidelines finds that the strongest evidence for efficacy is in major depression of at least moderate severity but states that efficacy for patients with mild depression, which presents frequently in primary care, is unproved. The draft guideline indicates that SSRIs are being overused in mildly depressed patients for whom the ratio between risk and benefits is unclear but may not be favourable.

A problem that is often overlooked is that the available randomised evidence does not provide reliable estimates of the costs and benefits of SSRI treatment in patients with varying levels of severity and baseline risk of suicide. For example, placebo controlled trials seem to show that SSRIs are only slightly better than placebo.⁵ Most trials are sponsored by the drug industry, which, if anything, might inflate estimates of efficacy.⁶ Publication bias favouring positive trials is widespread in SSRI versus placebo trials: many of the pivotal phase three trials investigating the efficacy of antidepressants have not been published.³,⁵ The potential for publication bias to cause serious bias and potential harm for many thousands of patients is well accepted, and it has been argued that legislation is now required.⁷ One obvious conclusion is that SSRIs do not work much better than placebo.

There are good reasons, however, to think that these trials do not tell the whole story and that the estimates of treatment effect that they produce may be seriously misleading. Head to head comparisons with the older tricyclic antidepressants indicate that, overall, SSRIs are of comparable efficacy⁸ and strategies to improve concordance with antidepressants usually lead to better outcomes.⁹

Regulators consider short term placebo controlled trials to be essential to demonstrate the efficacy of new antidepressants because of the high placebo response rate observed in the disorder.¹⁰ Yet placebo controlled trials are difficult to conduct because of the lack of clinical uncertainty about the efficacy of existing drugs for depression. Many UK ethical committees will not approve such trials, leading to a striking and unjustifiable lack of consistency between the national ethical and regulatory bodies. Because these trials are difficult to conduct, participants in placebo controlled trials tend to be highly selected. The increasing response to placebo over time indicates that this problem is getting worse,¹¹ with an increasing tendency for randomised trials to be carried out on people with relatively mild disorders with high spontaneous remission rates. This situation illustrates the dangers of relying entirely on trials conducted by industry to meet the limited needs of regulators in an important disease such as depression. Most registration trials are simply not designed to answer important clinical questions and to provide reliable and precise estimates of the benefits and risks of treatments. Independent trials are needed but, at least in the United Kingdom, non-commercial funding for trials is declining.¹²

Uncertainty therefore remains about the balance between benefits and harms of SSRIs. To us, however, it seems clear that the recent negative publicity about antidepressant drugs has fed into the routine stigmatisation and trivialisation of mental disorders. Doctors must not become reluctant to use antidepressant drugs in patients with clearly defined depressive disorders, but they should also monitor patients carefully in the first weeks of treatment.