Antidepressant Tapering Is Not Routine But Could Be

Discussion

In this primary care outpatient sample, only 8.9% of patients whose antidepressants were discontinued received prescriptions for the lowest available dose before stopping. The rate increased from 5% in 2014 but has plateaued around 10% in the past 4 years. Taper rates were not consistently predicted by antidepressants’ pharmacokinetics.

Note that although they may have had significant symptoms of withdrawal, it seems that most patients were able to stop taking antidepressants without tapering through low doses. On the other hand, thousands of patients report severe antidepressant withdrawal symptoms even with very small dose reductions.⁶ This suggests that some patients clearly require slow, small-decrement antidepressant tapers, whereas others can manage without using a lower dose or with no tapering at all.

Because it will dictate clinical practice, prescribers of antidepressants need to know how many patients need careful, slow tapering. A recent randomized trial suggests the answer is “less than half.”¹² In that trial’s taper process, doses were cut in half for a month; then that lowered dose was taken every other day for a month, then stopped. By 6 months, less than half of the discontinuation group had dropped out of the protocol, for example, to increase or resume their original dose. Thus, though some may have struggled in the process, the majority of participants successfully discontinued their antidepressant with this simple taper plan. Similarly, our data suggest that a majority of patients manage to discontinue their antidepressant without use of the lowest available doses.

However, many patients in the randomized trial, and in our clinics, might have experienced significant withdrawal symptoms even if these were not severe enough to lead to a change in strategy. A recent review found the incidence of withdrawal during antidepressant discontinuation averaged 56% of patients across 14 studies.⁴ While this study has aroused controversy,^5,13 it is nevertheless clear that a very large number of patients have an arduous course with current taper routines—or lack thereof. Worse yet, differentiating withdrawal from a relapse of the patient’s previous illness can be extremely difficult.¹⁴ Symptoms overlap, and time courses can be very similar.¹⁵

How then should clinicians discontinue antidepressants? Previous withdrawal problems or withdrawal symptoms between doses (particularly common with venlafaxine or paroxetine) help identify some patients who need small, slow decrements. But for the majority of patients, it seems that complex reductions are unnecessary.

Multiple taper strategies have been suggested, and recommended in practice guidelines,⁸ though few have been appropriately tested.¹⁶ A “hyperbolic” approach has been suggested, based on positron emission tomography scan receptor occupancy: large initial reductions followed by exponentially smaller decreases especially near zero.⁹ Decreasing doses can be facilitated with premade “tapering strips” containing progressively smaller doses,¹⁰ but these are not widely available.

We suggest a one-size-fits-all taper process, a variation of the hyperbolic approach based on our clinical experience, that balances the needs of patients at risk of severe withdrawal with the need for a simple clinical routine. Patients begin antidepressant discontinuation for different reasons, with different fears and different expectations. Yet they are alike in 1 crucial way: their first dose decrease must go well—no withdrawal effects and no worsening of mood—lest they immediately wish to go back to their original dose, from which it will be even harder to taper later.

Therefore, one should routinely begin tapering with the smallest decrement the patient can easily engineer. In most cases this will require a new prescription for the lowest available dose of the antidepressant. For patients who are anxious about potential relapse or withdrawal, cutting pills in half or even in quarters may be warranted.

After at least 2 weeks at the first step down, some patients will want to proceed with larger and/or more rapid reductions. If they encounter difficulties, they can slow down or even back up a step, but they will know from experience that small steps are manageable. At the other extreme, some patients will discover that even monthly small steps cause significant symptoms and debility. They can make smaller reductions, more slowly, using the liquid versions available for most antidepressants (notably not venlafaxine).

Approaching zero, many patients will need to slow down and proceed by again using the smallest decrements they can manage. Thus one could call this an “S-curve” variation of the hyperbolic approach: a small/slow step initially, potentially larger/faster steps from there, then small/slow steps again near zero. The developer of the hyperbolic taper acknowledges that an S-curve better describes his usual routine with patients (Mark Horowitz, personal communication, 2022).

(Note that even with this cautious approach, some patients will have severe withdrawal experiences. They may need the method used by many patients with severe withdrawal, reducing doses by as little as 10% of the previous month’s dose. This can take several years but has proved necessary for thousands of patients.¹⁷)