Are antidepressants overprescribed? No
BMJ 2013; 346 doi: https://doi.org/10.1136/bmj.f190 (Published 22 January 2013) Cite this as: BMJ 2013;346:f190
- Ian C Reid, professor of psychiatry
- i.reid{at}abdn.ac.uk
The notion that antidepressants are overprescribed is certainly popular and hardly new. There is profound suspicion of them: antidepressants are regularly caricatured in the media as an addictive emotional anaesthetic, peddled by thoughtless general practitioners as a matter of convenience, and taken by credulous dupes who seek “a pill for every ill.” Little wonder that decrying antidepressant prescription is such a sure-fire crowd pleaser for the press (for example, “Ministers act to wean Scotland off £55m-a-year antidepressant habit,” Scotsman, 6 December, 2006).
The reality is very different. Depressive disorder is a common, recurrent, debilitating, and potentially lethal illness. Psychiatric drugs, including antidepressants, have equivalent effectiveness to drugs in other branches of medicine (as detailed in a review of 94 meta-analyses comparing drug effect sizes in medical disease with drugs in psychiatric disorder1). Given recent demonstrations that depression is still under-recognised and undertreated,2 3 the claim that antidepressants are overprescribed needs careful consideration.
Reasons for the increase
Does the rising antidepressant prescription volume indicate overprescription? Many observers have assumed that increasing prescription must represent more patients diagnosed as having depression. Yet there has been no increase in incidence or prevalence; care seeking behaviour by patients; or identification by general practitioners4—rebutting the presumption that a creeping medicalisation of everyday distress is raising antidepressant use. The real reason is more mundane: a large descriptive study using the national general practice research database indicates that small but appropriate increases in the duration of antidepressant prescription—rather than more patients— have been driving the increase.5 The rise in volume simply represents gradually improving practice. Consider the arithmetic: many of the patients in this study were receiving treatment of inadequate duration (one month). Meeting current guidance (six months) should improve outcomes but would eventually increase antidepressant prescription volume sixfold, without any change in the numbers being treated.
Increased use in other conditions6 has compounded misunderstanding because the relevant statistics do not record diagnosis. In the last two years alone, Scottish national data show that the number of patients taking amitriptyline increased by 22%, accounting for nearly one third of all those receiving antidepressants.7 But this tricyclic is hardly ever prescribed for depression now; it is used instead for indications such as neuropathic pain.
The idea persists that GPs are handing out antidepressants “like sweeties.” We screened nearly 1000 general practice attenders in Grampian for depression and scrutinised the prescription decisions made by 33 GPs.3 Almost half of the depressed patients we identified were unrecognised and, contrary to popular stereotype, GPs were cautious and conservative in their prescribing for those that they did diagnose. We found only three patients for whom the indication was unclear. If only antibiotic or proton pump inhibitor prescription was so sparing. This finding helped persuade the Scottish government to withdraw a target to reduce prescribing by 10%.8
Practice supported by evidence
What about the widely reported story that antidepressants are no better than placebo? That arose from a meta-analysis of data from clinical trials submitted to the US Food and Drug Administration for the licensing of four of the new generation antidepressants.9 This was interpreted as showing that antidepressant drugs are no better than placebo except in severe depression. Sadly, demonstrations of methodological flaws and selective reporting10 suggest that the conclusions were “unjustified.”11 Another meta-analysis using complete longitudinal person level data from a large set of published and unpublished studies not only bolsters evidence for the efficacy of antidepressants but also suggests that baseline severity may not predict antidepressant response after all: milder cases seem to benefit, too.12 Contrary to current guidance, the question of efficacy in mild depression is not settled.
Media reports have claimed that limited availability of psychological therapy leads to inappropriate antidepressant prescription. Actually, there is no consistent relation between the availability of psychological therapies and antidepressant use—as shown by comparing the rates of antidepressant prescription with the numbers of talking therapists available across each of the primary care trusts in England.13 Furthermore, psychological therapies are, at best, as effective as antidepressant drugs, not superior to them; indeed, effect sizes for psychological therapy may be smaller.14 This is counterintuitive to both the public and politicians, many of whom assume that antidepressants cannot possibly be effective in the face of adversity. In fact, preceding adverse life events have little impact on response to antidepressants in depressive disorder.15
Antidepressants are but one element available in the treatment of depression, not a panacea. Like “talking treatments” (with which antidepressants are entirely compatible), they can have harmful side effects, and they certainly don’t help everyone with the disorder. But they are not overprescribed. Careless reportage has demonised them in the public eye, adding to the stigmatisation of mental illness, and erecting unnecessary barriers to effective care.
Notes
Cite this as: BMJ 2013;346:f190
Footnotes
I thank Isobel Cameron and David Christmas for helpful comments.
Competing interests: I have read and understood the BMJ Group policy on declaration of interests and declare I have received payment to deliver presentations at meetings organised by AstraZeneca and from Sanofi to attend an advisory board.
Provenance and peer review: Commissioned; not externally peer reviewed.