(The Lancet Psychiatry) – In their excellent manuscript, published in The Lancet Psychiatry, Mark Horowitz and David Taylor1 share their personal view on tapering of SSRI treatment to mitigate antidepressant withdrawal symptoms. In the Netherlands, this topic has been raising debate as well,2 which urged representatives of the Dutch college of General Practitioners, the Royal Dutch Pharmacists Association, the Dutch Association for Psychiatry, and the patient organisation MIND to develop multidisciplinary recommendations for the discontinuation of SSRIs and serotonin–norepinephrine reuptake inhibitors (SNRIs; appendix pp 1–15).
Independent of Horowitz and Taylor,1 but with identical reasoning, we proposed reducing doses of SSRIs or SNRIs hyperbolically, and we also advocated the strategy of mini-tapering (table; appendix pp 3–6).
|Citalopram (mg/day)||Escitalopram (mg/day)||Fluvoxamine (mg/day)||Paroxetine (mg/day)||Sertraline (mg/day)||Duloxetine (mg/day)||Venlafaxine (mg/day)|
Because the prevalence of antidepressant withdrawal symptoms is unknown1, 3 and not all patients require hyperbolic dose reductions to discontinue antidepressants, we identified three risk-factors that are consistently reported in the literature to indicate an increased risk of antidepressant withdrawal symptoms: (1) dosing above the minimal effective dose to reach efficacy, (2) antidepressant withdrawal symptoms when a dose was missed or during strategic treatment interruption, and (3) earlier failed attempts to discontinue the SSRI or SNRI. Although other risk factors have been proposed, these three factors might indicate a need to apply hyperbolic dose decreases instead of halving the minimal effective dose and stopping immediately therafter.
Instead of recommending reductions in steps of 1 month, which might be both very effective but also too conservative and time-consuming in a substantial number of patients who nevertheless require hyperbolic dose reductions, we propose initial steps of 1 week and observation of the emergence of antidepressant withdrawal symptoms. Such symptoms might either be bearable or, if not, would prompt a return to the dosage of a step earlier when no antidepressant withdrawal symptoms were noticed, followed by a slower discontinuation. This policy resembles the nomogram provided by Horowitz and Taylor,1 which is clinically very helpful.
Because of paucity of evidence to exactly guide clinical recommendations, we strongly advocate shared decision making and careful, repeated evaluation of the hyperbolic dosing steps. Advocating shared decision making could have been more explicit in the manuscript.
In 2018, Bockting and colleagues4 showed that the combination of preventive cognitive therapy (PCT) and maintenance antidepressants was superior (in terms of relapse prevention) to antidepressant maintenance therapy or PCT with discontinuation of antidepressants. In the same study, relapse rates were similar between antidepressant maintenance therapy and PCT with discontinuation of antidepressants. In a meta-analysis of four studies, Kuyken and colleagues5 showed that mindfulness-based cognitive therapy was superior to maintenance antidepressants. These publications emphasise the need for careful evaluation of the risk of recurrence or relapse before antidepressants are discontinued, and for initiating preventive psychotherapy to reduce risk of recurrence.
We agree with Horowitz and Taylor1 that empirical evidence regarding the discontinuation of antidepressants is urgently needed. Research should address: (1) the prevalence of antidepressant withdrawal symptoms in clinical populations discontinuing antidepressants, (2) systematic identification of risk factors for antidepressant withdrawal symptoms and their relevance being determined in prospective studies, and (3) the relative advantages of different dose-reduction regimens (eg, comparing mini-tapering and micro-tapering, and different timings of dose steps). In the ongoing OPERA project, the first two of these issues will be addressed, for the third, large collaborative randomised trials still need to be initiated.
For more on the OPERA project please see https://opera-project.nl