Increased prescription of antidepressants shows correction of inadequate duration of treatment of depression

Godlee draws attention to the Organisation for Economic Co-ordination and Development’s report that “the world’s rich nations have doubled their use of antidepressants in the past 10 years” and links this with Dowrick and Frances’ opinion piece “medicalising unhappiness” to suggest that antidepressants are being used inappropriately.1 2 3 Hence her advice: “Don’t keep taking the tablets.”

However, the raw prescription data reported tell us nothing about the appropriateness of prescribing. The rise in the UK is largely accounted for by the gradual correction of inadequate duration of treatment, not by increasing diagnosis of depression, which is static or falling.4 Doubling the duration of treatment doubles prescription volumes, without changing the number of patients being treated. Yet Dowrick and Frances say that overdiagnosis of depression is now more common than underdiagnosis. Unfortunately, the evidence cited does not show a change in diagnostic practice as implied,5 but rather that GPs are better at ruling out depressive disorder than detecting it (the “opposite” of overdiagnosis): they generate more “false positive” diagnoses simply because the prevalence of depression is modest. Mitchell aggregated studies from over two decades: his findings neither account for changes in prescription volume, nor show increasing medicalisation.

In a rapid response to our comments, Dowrick and Frances assert that “false negatives, better compliance, and the pain indication cannot account for the bulk of the remarkably increased use of antidepressants.”6 We agree. But their assumption that rising prescription rates are “more plausibly explained as a result of a loose diagnostic system, loosely applied, and pressured by extravagant Pharma marketing” cannot be sustained in the face of Moore and colleagues’ findings.4 Although we share concerns about the quality of diagnostic systems, many depressed people remain undetected and, when delivered, the duration of treatment is still inadequate.