Thread by @markhoro

Outline of the Cochrane study a bit for those scared of its length. Metro article: bit.ly/3tVfOah Lead author: the dauntless Ellen van Leuween. Co-author @tony_kendrick bit.ly/32KLJhj While there are hundreds of studies looking at starting antidepressants(1/n)

we found only 33 studies looking at stopping antidepressants, a therefore hugely neglected area 2/n

Most studies stopping antidepressants abruptly or in less than 4 weeks. This is not consistent with even the current lacklustre NICE guidance (stop over more than 4 weeks) and certainly nothing like the improved guidance from RCPsych bit.ly/3dHYUpL (3/n)

The main limitation of the studies was the failure to distinguish relapse from withdrawal, with most studies focusing on relapse -only one study measured withdrawal. As withdrawal can include low mood, anxiety, disturbed sleep and appetite (4/n)

All of which would register on depression scales used to detect relapse these would inflate the apparent rate of relapse found in the group who stopped antidepressants. Therefore there was very little to be gleaned from currently conducted studies (5/n)

None of the studies used changed pharmaceutical formulations allowing patients to reduce their dose below the commonly used therapeutic doses (6/n)

Implications for practice: clinicians should be aware of the distinction between withdrawal symptoms and relapse. Withdrawal symptoms can come on very quickly after stopping antidepressants, include symptoms not in the patient’s underlying condition (7/n)

Or other typical symptoms such as ‘shock-like’ or ‘electric zap’, dizziness, profound insomnia. (8/n)

Withdrawal symptoms are an indication to taper more slowly (perhaps after returning to a higher dose of medication to stabilise) not evidence that an underlying condition has returned (9/n)

Because relapse rates in studies are confounded by withdrawal effects it is not possible to be confident that current recommendations for continuation of treatment and maintenance for prevention of relapse are robust (10/n)

Implications for research: there is an urgent need for trials of stopping antidepressants that address confounding by withdrawal. These studies should have as their key outcome “successful discontinuation rate” and should involve a general population from primary care (11/n)

Future studies should “undertake long-term follow-up to evaluate person-centred outcomes such as quality of life, including return to work and daily and social activities” (12/n)

Although not the main aim of the review, one take away is that it is not possible to be confident in the current evidence recommending long-term treatment with antidepressants (13/n)

because this advice is based on studies in which patients are stopped off antidepressants and withdrawal is not taken into account (14/n)

@CochraneAus @CochraneUK @joannamoncrieff @AllenFrancesMD @RaniBora @RPurssey @jf_moore @Mental_Elf @eturnermd1 @Reduxreloaded

@WallisKatharine @NaudetFlorian