The Maudsley Deprescribing Guidelines: A Practical Implementation Review
The Maudsley Deprescribing Guidelines (Horowitz & Taylor, 2024) are the first formulary-grade manual dedicated to stopping psychotropic medication rather than starting it. They translate the pharmacology of receptor occupancy into concrete tapering schedules for antidepressants, benzodiazepines, z-drugs, gabapentinoids, and antipsychotics. For the prescriber, the central clinical shift is away from linear dose reductions and toward hyperbolic tapering — a method that addresses why so many patients stall at the final few milligrams.
Why the Maudsley Deprescribing Guidelines exist
The standard Maudsley Prescribing Guidelines have run to thirteen editions and are the most widely used psychotropic reference in the UK and much of the Commonwealth. Until 2024 they contained almost no operational guidance on discontinuation. The deprescribing volume fills that gap, and it does so because the evidence base on withdrawal had shifted decisively.
Two developments forced the issue. First, withdrawal reactions from antidepressants were re-characterised as common, sometimes severe, and frequently protracted — not the mild, self-limiting "discontinuation syndrome" described in older sponsor-funded literature. A 2019 systematic review by Davies & Read in Addictive Behaviors (pooled across 14 studies) reported that approximately 56% of patients stopping antidepressants experienced withdrawal symptoms and that 46% of those described them as severe. Second, the receptor-occupancy data explained why conventional tapers fail.
The guidelines are explicit that most withdrawal failures are iatrogenic — produced by reducing too fast and, critically, too linearly at the low end of the dose range. This reframing matters at the point of care: a patient who deteriorates during a taper is more often experiencing under-managed withdrawal than relapse.
The pharmacological core: receptor occupancy is hyperbolic, not linear
The single most important concept in the manual is the relationship between dose and target engagement. For SSRIs, serotonin transporter (SERT) occupancy follows a hyperbolic curve described by the law of mass action. Occupancy rises steeply at low doses and plateaus at clinical doses.
PET imaging data (Meyer et al., 2004, American Journal of Psychiatry) quantify this for citalopram:
| Citalopram dose | Approximate SERT occupancy | || | 2.5 mg | ~30% | | 5 mg | ~50% | | 10 mg | ~65% | | 20 mg | ~80% | | 40 mg | ~85% | | 60 mg | ~88% |
The clinical implication is direct. A reduction from 40 mg to 20 mg removes roughly 5 percentage points of SERT occupancy. A reduction from 5 mg to 0 mg removes roughly 50 percentage points. A linear taper — equal milligram steps — therefore delivers a trivial pharmacological change at the top of the range and a massive one at the bottom. Patients predictably tolerate the early reductions and then crash near the end, exactly when clinicians often assume the hard part is over.
Hyperbolic tapering corrects this by making each step produce an approximately equal reduction in receptor occupancy rather than an equal reduction in milligrams. In practice this means the dose decrements shrink as the dose falls, and the taper must extend below the smallest commercially available tablet.
What "equal occupancy steps" looks like in milligrams
If the goal is a consistent reduction in effect — the guidelines often target reductions of around 5–10% of receptor occupancy per step rather than 5–10% of dose — the milligram schedule becomes markedly non-linear. A worked sertraline example illustrates the shape (figures approximate, individualised to tolerance):
| Step | Sertraline dose | Approx. reduction in dose from prior step |
|---|---|---|
| 1 | 100 mg | — |
| 2 | 50 mg | 50% |
| 3 | 25 mg | 50% |
| 4 | 12.5 mg | 50% |
| 5 | 6 mg | ~50% |
| 6 | 3 mg | 50% |
| 7 | 1.5 mg | 50% |
| 8 | 0.8 mg | ~47% |
| 9 | 0 mg | final |
The proportional steps look aggressive on paper but produce roughly even pharmacological decrements. The final doses — sub-milligram for many drugs — are below anything a standard tablet provides, which is why formulation strategy is not optional.
Formulation strategies the guidelines endorse
Reaching 0.5 mg or 1 mg of a drug supplied as 25 mg or 50 mg tablets requires one of several approaches. The manual reviews each:
- Liquid formulations. Commercially available oral solutions (e.g., escitalopram, sertraline, paroxetine, fluoxetine, risperidone) allow precise small doses. This is the preferred method where a licensed liquid exists.
- Compounded suspensions or "tapering strips." Pharmacy-compounded liquids or pre-packaged daily-dose strips (widely used in the Netherlands) provide graded sub-tablet doses. The 2018 van Os & Groot work on tapering strips in Psychotherapy and Psychosomatics documented successful discontinuation in patients who had repeatedly failed conventional tapers.
- Dispersion in water. Many tablets can be dispersed in a fixed volume of water and a fraction taken. The guidelines note this is acceptable for tablets that disperse adequately but caution that distribution is not always uniform; the suspension must be stirred and used promptly.
- Tablet cutting. Useful only down to roughly a quarter tablet; below that, dose accuracy degrades and it cannot reach the sub-milligram range. Adequate for the top of a taper, inadequate for the tail.
The choice is governed by the drug's available formulations, not clinician preference, and the manual indexes this drug-by-drug.
Rate, monitoring, and the role of patient feedback
The guidelines do not prescribe a fixed calendar. Rate is titrated to the individual, with the symptom response to each reduction governing the timing of the next. The general structure:
- Reduce by a step that produces a tolerable, self-limiting set of symptoms.
- Hold until symptoms settle (commonly 2–4 weeks, sometimes longer).
- Only then proceed to the next reduction.
- If withdrawal is intolerable, return to the last well-tolerated dose, stabilise, and resume with smaller decrements.
This makes the taper patient-paced rather than protocol-paced. The manual is explicit that there is no minimum speed and that tapers extending over many months — occasionally longer for long-term benzodiazepine or antidepressant use — are legitimate and often necessary. Faster is not safer; the failure mode is almost always going too fast.
The Discontinuation Emergent Signs and Symptoms (DESS) checklist is recommended for structured monitoring, distinguishing the characteristic withdrawal profile (dizziness, "brain zaps," nausea, insomnia, heightened anxiety, sensory disturbance) from a returning mood or anxiety disorder.
Withdrawal versus relapse: the differential the guidelines emphasise
A recurring clinical error is misreading withdrawal as relapse and reinstating the drug indefinitely. The manual provides discriminating features:
| Feature | Withdrawal | Relapse |
|---|---|---|
| Onset after dose reduction | Days | Weeks to months |
| Somatic symptoms (dizziness, brain zaps, GI) | Prominent | Uncommon |
| Response to dose reinstatement | Rapid (often <48 h) | Slow (weeks) |
| Symptom novelty | New sensations not part of original illness | Recapitulates original episode |
| Trajectory | Improves as taper holds | Persists or worsens |
Rapid relief on reinstating a small dose is among the most useful diagnostic signs that the clinician is dealing with withdrawal, not recurrence. NICE guideline NG222 (2022) on medicines associated with dependence and withdrawal makes the same distinction and endorses gradual, individualised tapering for antidepressants and other dependence-forming drugs.
Beyond antidepressants
The manual extends the same hyperbolic logic across drug classes, each with its own occupancy curve and caveats.
- Benzodiazepines and z-drugs. Building on the Ashton Manual, the guidelines favour conversion to a long-half-life agent (diazepam) where appropriate, then hyperbolic reduction. GABA-A occupancy is likewise non-linear, so the tail requires the smallest steps.
- Antipsychotics. Dopamine D2 occupancy is hyperbolic; abrupt or rapid reduction risks dopaminergic rebound and supersensitivity psychosis. Horowitz, Jauhar, Natesan & Taylor (2021, Schizophrenia Bulletin) argued for tapers over months to years for long-term users, with final doses far below standard tablets.
- Gabapentinoids. Pregabalin and gabapentin produce a defined withdrawal syndrome; the guidelines apply proportional reductions and warn against the abrupt cessation common in primary care.
In each case the same two principles hold: steps proportional to dose (not fixed milligrams), and a tail that descends below the licensed minimum tablet.
Where the guidelines depart from older practice
The manual implicitly retires several widely taught heuristics that the occupancy data do not support.
The "10% every 2 weeks" rule. This linear shorthand fails for the same reason all linear tapers fail: a fixed percentage of the original dose is a fixed milligram step, which produces an accelerating occupancy change near zero. The guidelines instead frame reductions as a percentage of the current dose, recalculated at each step, so the absolute decrement shrinks as the patient descends.
The fluoxetine bridge. Substituting a long-half-life SSRI (fluoxetine, t½ of parent plus norfluoxetine metabolite spanning 4–16 days) to "self-taper" is presented in the manual as useful in selected cases but not a universal solution. Fluoxetine itself must still be tapered hyperbolically, and cross-taper tolerability varies. It does not eliminate the need to reach sub-clinical doses; it changes the kinetics of how the patient gets there.
"It's your depression returning." The guidelines are pointed that reflexively attributing post-reduction symptoms to relapse — and reinstating indefinitely — is a frequent error. The temporal and somatic profile usually distinguishes the two, and the reinstatement-response test settles most ambiguous cases within 48 hours.
Protracted withdrawal is real. A minority of patients, particularly after years of treatment, experience symptoms lasting months after the last dose. The manual treats this as a recognised phenomenon requiring an even slower tail, not evidence of an underlying disorder. Acknowledging it explicitly to the patient improves adherence to a slow schedule.
Implementing this in a real clinic
Translating the manual into practice does not require specialist infrastructure, but it does require front-loaded planning.
- Establish the indication for stopping and the patient's goals. Hyperbolic tapering is slow by design; the patient must understand the timescale (often 6–12 months, sometimes longer) before starting.
- Map the formulation route to zero at the outset. Confirm whether a licensed liquid exists, whether the local pharmacy will compound, and whether tablets disperse. This determines the floor the patient can reach.
- Write the schedule as percentages, not fixed milligrams. Specify the proportional step and the minimum hold interval, with an explicit instruction to pause or step back up if symptoms are intolerable.
- Give the patient a symptom log. A DESS-style checklist, reviewed at each contact, both guides the pace and documents the withdrawal-versus-relapse picture.
- Schedule follow-up contingent on the taper, not the calendar. Reductions in the difficult low-dose phase warrant closer contact than the early steps.
Shared decision-making is integral. The patient sets the pace within a clinician-defined safety envelope; documenting that the slow timescale was discussed and agreed pre-empts the pressure to rush that derails most attempts.
Clinical pearls
- Plan the tail before the top. The hardest part of any taper is the final few milligrams. Decide the formulation strategy (liquid, compounded suspension, dispersion) at the outset, not when the patient hits the lowest tablet.
- Make each step proportional, not fixed. Reducing by the same milligram amount each time guarantees an accelerating pharmacological effect near zero. Reduce by a roughly constant percentage of the current dose, shrinking the absolute steps as the dose falls.
- Use reinstatement response as a diagnostic test. Rapid improvement within 24–48 hours of returning to the prior dose points to withdrawal; a slow response over weeks points to relapse. Do not commit a patient to indefinite treatment on the basis of early post-reduction symptoms.
- Let symptoms set the pace. Hold at each step until symptoms settle before the next reduction. There is no minimum rate; the only reliable error is going too fast.
- Document with DESS or an equivalent. Structured symptom tracking separates the somatic withdrawal profile from a returning mood disorder and supports the patient through the difficult low-dose phase.
- Match the method to the drug's formulations. Whether a clean hyperbolic taper is achievable depends on whether a liquid or compounding option exists. Check this drug-by-drug rather than assuming tablet-cutting will suffice.
References
- Horowitz MA, Taylor D. The Maudsley Deprescribing Guidelines. Wiley, 2024.
- Davies J, Read J. A systematic review into the incidence, severity and duration of antidepressant withdrawal effects. Addictive Behaviors, 2019.
- Meyer JH, et al. Serotonin transporter occupancy of five SSRIs at different doses: a PET study. American Journal of Psychiatry, 2004.
- Horowitz MA, Taylor D. Tapering of SSRI treatment to mitigate withdrawal symptoms. Lancet Psychiatry, 2019.
- Horowitz MA, Jauhar S, Natesan S, Taylor D. A method for tapering antipsychotic treatment that may minimize the risk of relapse. Schizophrenia Bulletin, 2021.
- Groot PC, van Os J. Antidepressant tapering strips to help people come off medication more safely. Psychotherapy and Psychosomatics, 2018.
- National Institute for Health and Care Excellence. Medicines associated with dependence or withdrawal symptoms (NG222). 2022.
For more clinician resources on safe deprescribing and tapering, visit tapermeds.com.