Informed Consent for Deprescribing: What to Document Before Tapering

Informed consent for stopping a medication receives less attention than consent for starting one, yet the medico-legal and clinical stakes are comparable. Withdrawal phenomena, relapse risk, and the possibility of protracted symptoms after discontinuation of antidepressants, benzodiazepines, antipsychotics, gabapentinoids, and stimulants are now well-characterized in the literature, and a prescriber who initiates a taper without a documented discussion of these risks is exposed both clinically and legally. This article outlines what to elicit, what to disclose, and what to document before the first dose reduction.

Why consent for deprescribing requires its own framework

Consent for prescribing typically addresses indication, expected benefit, common adverse effects, and alternatives. Consent for deprescribing must additionally address the pharmacology of withdrawal, the difficulty of distinguishing withdrawal from relapse, the time horizon of the taper, and the possibility — supported by Horowitz and Taylor (2019) in Lancet Psychiatry, the Maudsley Deprescribing Guidelines (Horowitz & Taylor, 2024), and an expanding base of patient-reported outcome data — that some patients develop protracted withdrawal lasting months or, in a minority, years.

The 2019 Royal College of Psychiatrists position statement on antidepressant withdrawal explicitly acknowledged that withdrawal effects can be "severe and prolonged" in some patients, reversing prior NICE language. The 2022 NICE guideline NG215 on medicines associated with dependence or withdrawal symptoms (covering opioids, benzodiazepines, gabapentinoids, Z-drugs, and antidepressants) requires shared decision-making and documented discussion before tapering. Documentation that meets these standards is no longer optional in jurisdictions that have adopted equivalent guidance.

The four domains to cover before any dose reduction

A consent conversation that holds up clinically and medico-legally addresses four domains. Each should appear in the chart note in some form.

1. Indication review and the rationale for stopping

Before discussing taper mechanics, confirm and document why discontinuation is being considered. Common rationales include:

• Resolution of the original indication (e.g., a single depressive episode now in sustained remission >12 months on antidepressant)
• Adverse effects outweighing benefit (sexual dysfunction, weight gain, cognitive blunting, anhedonia, QTc prolongation)
• Drug interactions or new comorbidities
• Patient preference after weighing risks and benefits
• Pregnancy planning or pregnancy
• De-prescribing of polypharmacy in older adults (Beers Criteria 2023, STOPP/START v2)

Document the original indication, the duration of treatment, prior taper attempts, and the trigger for the current discussion. The phrase "patient requests discontinuation" alone is insufficient — pair it with a clinician assessment of whether the request is clinically reasonable.

2. Relapse risk

Quantify it where possible. For major depressive disorder, the 2003 Geddes meta-analysis in The Lancet found a relapse rate of 41% in patients discontinuing antidepressants over 1–3 years versus 18% on continued treatment. For schizophrenia, the Wunderink 7-year follow-up (2013, JAMA Psychiatry) showed higher relapse but better functional outcomes in a dose-reduction arm — a finding the consent conversation should reflect rather than gloss over. For benzodiazepine-treated panic disorder or chronic insomnia, relapse of the underlying condition is common but often confounded by withdrawal.

Document the relapse rate the patient was told, the timeframe, and any individualized risk factors (number of prior episodes, residual symptoms, family history, current psychosocial stressors).

3. Withdrawal risk

This is the domain where consent most often fails. The Davies and Read (2019) systematic review in Addictive Behaviors found that 56% of patients discontinuing antidepressants experience withdrawal symptoms, with approximately half rating them severe; 25–50% of episodes meet criteria for protracted duration. The figures are higher still for paroxetine, venlafaxine, and short-half-life agents.

The conversation should cover:

• The expected symptom profile for the specific drug class (e.g., FINISH mnemonic for SSRIs/SNRIs: Flu-like, Insomnia, Nausea, Imbalance, Sensory disturbance, Hyperarousal)
• Onset (typically within 2–4 days for short-half-life agents, 1–2 weeks for fluoxetine)
• Expected duration with appropriate hyperbolic tapering versus linear or abrupt discontinuation
• The possibility of protracted withdrawal — a low but non-zero probability that should be named, not omitted
• For benzodiazepines specifically: the risk of withdrawal seizures with abrupt cessation, autonomic instability, and the Ashton Manual recommendation of crossover to diazepam for short-acting agents
• For antipsychotics: supersensitivity psychosis, tardive dyskinesia unmasking, and cholinergic rebound

Documenting that the patient was informed of the possibility of protracted withdrawal is the single most important defensive entry in the chart. Patients who develop prolonged symptoms and were never warned often pursue complaints or claims; patients who were warned and chose to proceed almost never do.

4. The taper plan and decision points

Specify the schedule in writing. A consent discussion that ends with "we'll go slow" is not consent. The schedule should include:

• Starting dose and target (off, or a maintenance reduction)
• Reduction increment as a proportion of the current dose, decelerating as the dose approaches zero (the Maudsley guidelines and Horowitz & Taylor describe hyperbolic tapering grounded in the curvilinear relationship between dose and serotonin transporter occupancy)
• The formulation strategy (compounded liquid, tapering strips, manufacturer-approved bead or tablet division, switching to a long-half-life agent)
• Pre-defined hold or reverse criteria (e.g., "if symptoms are CGI-S ≥4 for >1 week, hold; if >2 weeks, return to previous dose")
• Monitoring schedule (in-person or telehealth visits, validated scales such as DESS for SSRI withdrawal, BWSQ for benzodiazepines)
• Escape conditions (resumption of full dose, referral, hospitalization triggers)

Suggested clinician language

Patient-facing scripts shorten the consultation and make the documentation easier. Adapt these to the specific agent and patient.

Framing the conversation:

"Before we lower the dose, I want to walk you through what we know about coming off this medication, what we don't know, and what the plan looks like. The decision is yours, but I want it to be informed. I'll write the key points in your chart so we both have a record."

Disclosing withdrawal risk for an SSRI/SNRI:

"About half of people who reduce or stop this class of medication experience some withdrawal symptoms — most often dizziness, nausea, sleep disturbance, electric-shock sensations, and emotional changes. For most patients these last a few weeks once the dose stabilizes. In a smaller group, they last longer — months in some cases, and in rare instances longer than that. Reducing the dose gradually does not eliminate the risk, but it lowers it substantially compared with abrupt discontinuation."

Disclosing withdrawal risk for a benzodiazepine:

"Benzodiazepines need to come down slowly because abrupt reduction can cause severe withdrawal, including seizures. The reduction is in small steps with the option to hold whenever symptoms become difficult, and the timeline is typically months, not weeks. If at any point you don't feel safe with the pace, the plan is to slow down or pause. This is not a measure of willpower — it is pharmacology."

Disclosing relapse risk:

"Stopping the medication carries a real chance that the underlying condition returns. For your situation, with [N] prior episodes and [current status], the risk over the next year is approximately [X]%. There is no way to predict who will relapse, but early signs are something to monitor for, and treatment can be restarted quickly if needed."

Closing the consent conversation:

"Given what I've described, the options are to proceed, to wait, or to look at alternatives such as a dose reduction without full discontinuation. There is no wrong answer, and this can be revisited at any visit."

Documentation template

A defensible chart note for the initial deprescribing consent visit covers the following fields. The exact format varies by EMR; the content should not.

| Field | Example entry | || | Indication and treatment duration | MDD, single episode, full remission >18 months on sertraline 100 mg | | Rationale for taper | Patient request; sustained remission; sexual dysfunction | | Prior taper attempts | None / one attempt 2024, abrupt cessation, withdrawal at 5 days, resumed | | Relapse risk discussed | ~40% over 1–3 years (Geddes 2003) cited; individualized to patient | | Withdrawal risk discussed | ~50% incidence; FINISH symptoms named; protracted withdrawal acknowledged | | Specific risks for this agent | Sertraline t½ ~26 h; moderate withdrawal liability; no active metabolite | | Taper schedule agreed | Hyperbolic schedule using compounded liquid; written plan provided | | Hold/reverse criteria | DESS ≥10 sustained 1 week → hold; ≥2 weeks → revert to prior dose | | Monitoring plan | Telehealth check-in 2 weeks after each reduction; PHQ-9 q4w | | Patient questions and decision | Asked about duration; chose to proceed | | Written information provided | Maudsley patient-facing taper sheet; tapermeds.com | | Capacity assessment | Capacity intact; understood and repeated back risks |

The "repeated back risks" entry is small but disproportionately useful. A single sentence — "patient repeated back the key risks of relapse and withdrawal in their own words" — is the cleanest documentation that consent was understood, not merely delivered.

Special populations

Older adults

For patients ≥65, the consent discussion should include falls risk reduction as a benefit (particularly for benzodiazepines, Z-drugs, and anticholinergics on the Beers list), cognitive effects, and the practical challenge of polypharmacy taper sequencing. Document the Drug Burden Index or anticholinergic burden score where relevant.

Pregnancy and lactation

Consent must address both the risks of continuation (e.g., neonatal adaptation syndrome with SSRIs in late pregnancy) and the risks of discontinuation (relapse rates of 68% in pregnant women who discontinued antidepressants in the Cohen 2006 JAMA cohort, versus 26% in those who continued). The conversation is rarely simple; a single visit is usually inadequate, and joint review with obstetrics or perinatal psychiatry should be documented.

Patients with prior severe withdrawal

A prior bad taper changes the risk calculus. Document the prior experience in detail — agent, schedule, symptoms, duration — and the modifications made for this attempt (gentler initial reductions, liquid formulation, longer intervals between reductions, or cross-titration to a longer-half-life agent). Patients with prior protracted withdrawal warrant the most conservative end of the recommended range and a longer interval between reductions.

Capacity-limited patients

For patients with cognitive impairment, severe ongoing psychiatric illness, or under involuntary treatment orders, consent procedures follow jurisdictional substitute decision-making law. The chart should reflect the capacity assessment, the substitute decision-maker's identity if applicable, and the rationale for the deprescribing decision in that framework.

Common documentation failures

The most frequent gaps in deprescribing consent notes, drawn from medico-legal review patterns:

• Withdrawal risk mentioned generically but the possibility of protracted symptoms not named
• Taper schedule absent from the note ("agreed to taper" with no formulation strategy or monitoring intervals)
• No pre-defined hold or reverse criteria — leaving the prescriber to improvise when symptoms emerge
• No record of written patient information provided
• Capacity not addressed when there is reason to question it (active depressive episode, cognitive changes)
• No documentation that the patient could ask questions and did

A consent note that addresses each of these in one or two sentences is rarely longer than 250 words and is substantially more defensible than a longer narrative that omits them.

Reassessing consent during the taper

Consent at the start of a taper does not cover every decision made six or twelve months later. The withdrawal experience often intensifies in the final portion of the dose range, where the receptor occupancy curve is steepest. A brief re-consent at major decision points — entering the lower dose range, switching formulation, holding for symptoms, or reversing — should be documented in the same structure: what was discussed, what was decided, and what the patient understood. This protects both the prescriber and the patient by ensuring that course corrections are recorded as deliberate clinical decisions rather than improvisation.

The same applies if new evidence emerges that changes the risk profile of the agent (for example, a regulatory update or a meta-analysis revising the withdrawal incidence figures). A one-line note stating that the updated information was shared and the patient elected to continue is sufficient.

Clinical pearls

• Quantify risks in the note. "About half of patients" is better than "some patients"; a cited figure with a source is better still.
• Explicitly name the possibility of protracted withdrawal for antidepressants, benzodiazepines, antipsychotics, and gabapentinoids. Omission is the single highest-yield failure mode.
• Write the taper schedule in the chart with the formulation strategy, monitoring intervals, and pre-defined decision points. A schedule the patient can read back is a schedule that has been consented to.
• Provide written patient-facing material and document that you did. A consent discussion with no take-home reference rarely survives a one-week recall test.
• Reassess consent at each reduction. Brief re-consent at major decision points — particularly entering the lower dose range, where withdrawal often intensifies — is appropriate.
• Document the patient repeating back the key risks. One sentence; outsized medico-legal value.

---

For more clinician resources on safe deprescribing and tapering, visit tapermeds.com.