Benzodiazepine tapering: decision and pace

A woman in her early seventies returns to the outpatient clinic to renew her usual prescription. She has taken the same sleeping pill for more than ten years, since her husband died. She fell in the bathroom two months ago, with no fracture, and her daughter comments that her mother has been forgetful. The prescription is ready on the screen; signing it would be enough. You hesitate: proposing a reduction now may sound like an accusation, and leaving things as they are may be the riskiest decision of the visit. This text uses benzodiazepine tapering as a setting for clinical reasoning: when to propose a reduction, how to negotiate the pace with the person, how to separate withdrawal from the return of the underlying symptom, and which biases push the decision one way or the other. The scene and the examples throughout the text are fictional and were written for teaching.

NEXO is a daily medical diagnosis game: one clinical case a day, up to six progressive clues, and ICD-10 proximity feedback across five levels.

The text also avoids being a dosing protocol. Tapering schedules, equivalences and timelines follow the guidelines and protocols of each service, under supervision. Here the focus is on how to think: which questions to ask before deciding, how to interpret what happens during the reduction, and how to keep the person at the center of the decision. If you play the case of the day on NEXO, you will recognize the movement: each new clue changes the weight of the hypotheses, and the most recent clue rarely explains everything on its own.

Key points

• Long-term benzodiazepine use calls for periodic reassessment of risks and benefits, and gradual reduction is a shared decision, not an order.
• Abrupt discontinuation in someone with physical dependence can cause severe withdrawal, including seizures; the pace of reduction is a central part of safety.
• During reduction, three phenomena get confused: withdrawal, rebound and the return of the underlying symptom. Telling them apart changes management.
• Prescribing inertia, the "dependent" label and attributing any complaint to withdrawal are frequent biases in this setting.
• Clear information for the person and psychosocial support increase the chance of success, according to trials and systematic reviews.

Why automatic renewal is a clinical decision

Renewing a prescription looks like an administrative act, but each renewal is a decision to keep a treatment going. With long-term benzodiazepine use, that decision weighs more as the years pass, because the benefits tend to shrink and the risks tend to grow. Malcolm Lader's review of benzodiazepine harms describes sedation, cognitive and psychomotor impairment, difficulty with complex tasks such as driving, as well as dependence and withdrawal (Lader, 2014). In older adults, the American Geriatrics Society Beers Criteria list benzodiazepines among potentially inappropriate medications, because of the risk of cognitive impairment, delirium, falls, fractures and traffic accidents (AGS, 2023).

The name of the phenomenon that keeps the prescription going is prescribing inertia: the tendency to continue what is already underway because changing it takes time, conversation and the risk of conflict. Inertia rests on well-known biases. Status quo bias favors the option that requires no action. Omission bias makes the harm caused by not acting seem smaller than the harm caused by acting. Pat Croskerry describes these shortcuts among the cognitive errors that affect clinical decisions (Croskerry, 2003). If you want to review how useful shortcuts become a trap, the post on examples of heuristics in clinical reasoning organizes the main ones.

The cascade that starts with one pill

Rochon and Gurwitz described the prescribing cascade: an adverse effect is read as a new problem and gets another medication, which brings new effects (Rochon and Gurwitz, 1997). With benzodiazepines, the cascade shows up in several ways. Daytime drowsiness becomes a complaint of fatigue. Unsteadiness becomes a workup for dizziness. Forgetfulness becomes a suspicion of dementia. At each step, the reasoning needs to include a simple question: does any medication in use explain this finding?

Reassessing is different from withdrawing

The 2025 joint guideline on benzodiazepine reduction, coordinated by the American Society of Addiction Medicine with nine other societies, recommends continuous assessment of risks and benefits, shared decision-making with the person, tailoring the reduction to each case, and adjusting the pace according to the response (Brunner et al., 2025). Reassessment may end in maintaining, reducing slowly, or changing the treatment strategy for the underlying symptom. The mistake is to stop asking.

When to propose reduction: clues that change the weight of the decision

The question "should I propose tapering?" resembles any diagnostic decision: there are clues that increase and clues that decrease the probability that reduction will bring benefit. Organizing these clues avoids two opposite failures. One is never bringing up the subject. The other is proposing withdrawal to everyone in the same way, without looking at the person.

Clues that favor the conversation about reduction:

• Falls, instability, or recent fracture, especially in older people.
• Cognitive complaints, slowness, or daytime drowsiness.
• Concomitant use of opioids, alcohol, or other central nervous system depressants.
• An original indication that no longer exists, such as a bereavement from many years ago.
• Doses that increase over time or prescriptions sought from multiple professionals.
• The person's own desire to reduce, often kept hidden for fear of worsening.

The combination with opioids deserves emphasis. In 2020, the FDA required a boxed warning on the label of the entire benzodiazepine class about abuse, misuse, dependence, and withdrawal reactions, and reinforced the risk of use with opioids (FDA, 2020). The post about opioid overdose and the naloxone decision discusses what happens when depressants add up on shift.

Clues that call for caution before reducing

Some situations require more preparation before starting: a history of withdrawal seizure, use of high doses, active alcohol use disorder, a current psychiatric crisis, or a very unstable life moment. In these situations, the reasoning usually leads to planning with more support, not to giving up. The 2025 joint guideline includes assessment of these factors in the decision about the pace and setting of reduction (Brunner et al., 2025).

The fictional example, again

In the opening scene, there is a recent fall, a cognitive complaint, and an indication that arose from an old bereavement. The clues point to talking about reduction. Questions remain: does she use alcohol? Does she take other sedatives? Has she tried to stop, and what happened? What does she think about the pill? These answers define the pace, not just the decision.

How to open the conversation without a label

The way tapering is proposed changes the person's response. Someone who hears "you are dependent" tends to become defensive. Someone who hears "this medicine may be linked to your falls, and I wanted to think with you about a safe way to reduce" has room to participate. Kelly and Westerhoff showed, in a randomized study with mental health professionals, that describing the same person as a "substance abuser" instead of a "person with a substance use disorder" increased blame attribution and preference for punitive measures (Kelly and Westerhoff, 2010). The post about stigma and language shows how the label changes conduct.

Many people on long-term use started treatment with a medical prescription and followed the guidance they received. Physical dependence, in this case, is an expected consequence of continuous use, not a character flaw. Calling this "addiction" pushes the person away from the conversation and removes their willingness to recount previous attempts.

What the evidence says about informing the person

A randomized cluster trial, conducted in community pharmacies in Canada with older adults on long-term benzodiazepine use, tested an educational leaflet sent directly to the patient, with the risks of the medication and a tapering schedule. The group that received the material stopped use more often than the usual care group, and many participants started the conversation with the doctor or pharmacist on their own (Tannenbaum et al., 2014). The point for reasoning is that clear information can open shared decision-making, not replace it.

Motivational conversation helps when there is ambivalence: the person wants to reduce falls and fears insomnia. The post on motivational interviewing for students and residents explains how to explore both sides without pressure.

Phrases that often help

• "How do you feel about this medicine today, compared with the beginning?"
• "What worries you when you think about reducing?"
• "We can go slowly, with pauses, and adjust together if something gets worse."
• "If at some point it gets hard, we review the agreement, without blame."

Pace of reduction: shared decision, not a formula

The pace is the heart of tapering. Two meta-analyses show that gradual reduction beats usual care. Parr et al. found better discontinuation rates with gradual dose reduction and with brief interventions, compared with routine care, and an additional gain when gradual reduction was combined with a psychological intervention (Parr et al., 2009). Voshaar et al. described favorable results for minimal interventions, such as a letter with guidance, and for systematic discontinuation programs led by a doctor or psychologist (Voshaar et al., 2006).

The 2025 joint guideline recommends adapting the reduction to each person and adjusting the pace according to the response, and advises against abrupt discontinuation in those who likely have physical dependence (Brunner et al., 2025). In reasoning practice, this means treating the pace as a hypothesis to be tested: start with one step, observe the response and decide the next step with the person.

Switching to a long half-life

One strategy described in the literature is to switch from a short half-life benzodiazepine to a long half-life one before reducing, to smooth the concentration swings between doses. The Ashton Manual, written by psychopharmacologist Heather Ashton from the experience of a withdrawal clinic in the United Kingdom, popularized this method and the equivalence tables used by many people who taper (Ashton, Ashton Manual). The 2025 joint guideline allows considering a switch to an equivalent dose of a longer-acting benzodiazepine in those without contraindication, as a conditional recommendation based on clinical consensus, and calls for attention to liver function and to other medications in use (Brunner et al., 2025). The same guideline considers the switch less suitable in older adults, because of the risk of accumulation and sedation (Brunner et al., 2025). For the woman in the scene, therefore, the switch is the least indicated option, and the choice follows the case and the local protocol.

What the evidence shows about adjuvant medications

A Cochrane review on pharmacological interventions to support discontinuation in chronic users found low-quality evidence, insufficient to recommend an adjuvant medication routinely (Baandrup et al., 2018). Another Cochrane review, on psychosocial interventions, found benefit from cognitive behavioral therapy combined with gradual reduction in the short term, with less clear evidence in longer follow-up (Darker et al., 2015). For reasoning, the message is that psychological support, especially for insomnia and anxiety, is part of care from the start.

Withdrawal, rebound or return of the underlying symptom

During the taper, the person comes back saying they slept badly, that they are anxious or that their heart is racing. The temptation is to attribute everything to withdrawal and go back to the previous dose, or to attribute everything to the "psychiatric picture" and keep the pace. Both readings can be wrong. The literature usually separates three phenomena (Soyka, 2017; Lader, 2014):

Withdrawal. Symptoms that may be new or different from the original picture, such as muscle spasms and tension and heightened perception, as well as anxiety and insomnia. They appear after the reduction and tend to improve over time.

Rebound. Return of the original symptom with greater intensity than before treatment, usually transient and soon after the reduction.

Return of the underlying symptom. The problem that led to the prescription comes back, with features similar to those from before, and tends to persist. It calls for its own treatment, not just more benzodiazepine.

Clue	Withdrawal	Rebound	Return of the underlying symptom
Relation to the reduction	Starts after the reduction	Starts soon after the reduction	May appear weeks later
Type of symptom	New symptoms, such as muscle tension and sensory hypersensitivity, as well as anxiety and insomnia that overlap with the original picture	Same symptom, more intense	Same symptom, with a pattern similar to the original
Course	Tends to improve over time, although in some people the symptoms persist for months	Transient	Persists
What usually helps	Pause or slow the reduction	Observation and pace adjustment	Treat the underlying problem, such as therapy for insomnia
Most common bias	Being read as "weakness" in the person	Being read as proof that the drug is indispensable	Being read as eternal withdrawal

In practice, the separation is often difficult, and withdrawal complaints are sometimes mistakenly attributed to the return of the original picture (Brunner and colleagues, 2025). The table is a support for reasoning, not a classification rule: the three phenomena mix, and the response to time and to pace adjustment is usually the most useful clue.

The bias of attributing everything to withdrawal

When the person is tapering, any new complaint risks being read as withdrawal. It is the same mechanism as diagnostic overshadowing: a strong label covers other hypotheses. Palpitations may be withdrawal, but they may be arrhythmia, hyperthyroidism or stimulant use. Confusion may be withdrawal, but it may be infection, hyponatremia or another medication. The post on CIWA-Ar shows how benzodiazepine withdrawal mimics alcohol withdrawal, and the same logic applies: the scale measures symptoms, not their cause.

Severe withdrawal: when the pace becomes a safety issue

Benzodiazepine withdrawal can, less often, progress with seizures or psychosis (Lader, 2014; Soyka, 2017), and the risk grows when the stop is abrupt in someone with physical dependence. The 2025 joint guideline places the recommendation against abrupt discontinuation among its central messages (Brunner and colleagues, 2025). This situation appears in contexts that reasoning needs to anticipate:

• Hospital admission where the usual medication disappears from the prescription by oversight.
• Prison, moving to another city or loss of access to the prescriber.
• A person who decides to stop alone after reading about the risks.
• Use of benzodiazepines obtained without a prescription, with irregular doses.

That is why medication reconciliation on admission is a safety step. Asking "what medicine do you take to sleep or for your nerves?" finds the benzodiazepine that the formal list forgot. On shift, a seizure or delirium without a clear clinical cause in someone who uses a benzodiazepine calls for the question about recent discontinuation.

Alcohol and benzodiazepine in the same patient

When the person uses alcohol and a benzodiazepine, withdrawal from one can mask withdrawal from the other, and reasoning needs to consider both. The post on alcohol beyond withdrawal discusses the risks of labeling the intoxicated patient too early, and the same care applies here.

The biases that get in the way of tapering

It is worth gathering the biases in one place, because they appear at different moments of follow-up:

1. Status quo and omission. Keeping the prescription seems neutral, although it is a choice with risks.
2. Label of "dependent". Treats the expected physical dependence as a moral problem and pushes the person away.
3. Anchoring on abstinence. Every new complaint becomes abstinence, and other causes stop being investigated.
4. Rush to finish. The pace is sped up to meet a schedule, not to follow the response.
5. Affective bias. Irritation with prescription requests or with difficult consultations enters the decision without being noticed.

The post on affective bias explains how the professional's emotions change reasoning, and the topic comes up often in consultations about controlled medications. The post on calibration of confidence in the diagnosis offers a useful tool for each reassessment: record how much you trust your hypothesis and check it later.

When not to reduce quickly

Some people need a very slow pace, with long pauses. Others choose to keep using after an honest conversation about risks. Respecting that choice, reviewing it from time to time and reducing other risks, such as combining with opioids, is also a valid clinical decision. The post on harm reduction beyond all or nothing develops this logic.

How NEXO helps train this reasoning

In NEXO, the case of the day is fictional, with a real clinical presentation and up to six progressive clues. You propose a diagnosis by searching for the ICD-10 code and receive proximity feedback in five levels, which shows whether the guess landed on the exact code, the same category, the same group, the same chapter or a different chapter. This format connects with the theme of this text: a medication in use is a clue that may explain the picture, and the feedback shows whether the explanation you chose was close to the answer.

To use the case of the day as reasoning training about medications, try three habits. Before each guess, ask whether any medication in use would explain the clue. When an explanation seems obvious, write down the second most likely hypothesis. After the result, note which clue you overvalued.

Three exercises with cases

Review the medication list. In each case, note the medications mentioned and ask which of them could cause the main symptom.

Separate the three phenomena. For a fictional reduction case, write what favors abstinence, rebound and return of the underlying symptom.

Rewrite the conversation. Take a chart sentence with a label and rewrite it in person-centered language.

The post on learning based on clinical cases shows why training with cases helps fix this type of decision.

The limits on using the game are described on the NEXO economy page. If you research teaching of clinical reasoning, the page for researchers explains how to get in touch. Professionals who write or review cases find information under curation and under creators, and institutions can learn about the proposal on the partnerships page.

Frequently asked questions

Should every person on long-term benzodiazepine use taper off?

The decision is individual. The 2025 joint guideline recommends reassessing risks and benefits continuously and deciding together with the person (Brunner et al., 2025). In older people, in those with falls or in those using other depressants, the risks usually weigh more.

What is the ideal pace of reduction?

The pace is adjusted to each person's response, with small steps and pauses when needed. Specific schedules follow the guidelines and the service's protocols, under supervision.

How do you tell abstinence from the return of anxiety?

Abstinence usually brings new symptoms and improves over time; the return of the underlying symptom resembles the original picture and persists. The response to adjusting the pace is a useful clue (Lader, 2014).

Is it worth switching to a long half-life benzodiazepine?

Switching is an option described in the Ashton Manual and accepted by the 2025 guideline as a conditional recommendation, more suitable in adults without contraindication and less suitable in older people (Brunner et al., 2025). The choice depends on the case, the age and the local protocol.

What should be done if the person wants to stop alone, all at once?

Explain the risk of abrupt interruption, including seizure, and offer a gradual reduction with follow-up. If there are signs of severe abstinence, seek the necessary help at the emergency service.

Does therapy help with tapering?

Help. Reviews show benefit from cognitive behavioral therapy combined with gradual reduction, especially for insomnia and anxiety (Darker et al., 2015).

Final checklist

• I reviewed the original indication and the current risks of the benzodiazepine.
• I asked about falls, memory, alcohol, opioids and other sedatives.
• I proposed the reduction without labels and listened to what the person fears.
• I agreed on an initial pace and a date to reassess the response.
• I separated withdrawal, rebound and return of the underlying symptom before changing the pace.
• I offered psychological support for the underlying symptom.
• I recorded the regimen so that another professional can continue.

Want to practice the reasoning with progressive clues? Play the case of the day.

References

1. Ashton CH. Benzodiazepines: How They Work and How to Withdraw (Ashton Manual). benzo.org.uk.
2. American Geriatrics Society Beers Criteria Update Expert Panel. American Geriatrics Society 2023 updated AGS Beers Criteria for potentially inappropriate medication use in older adults. J Am Geriatr Soc. 2023. PMID 37139824.
3. Baandrup L, Ebdrup BH, Rasmussen JØ, et al. Pharmacological interventions for benzodiazepine discontinuation in chronic benzodiazepine users. Cochrane Database Syst Rev. 2018. PMID 29543325.
4. Brunner E, Chen CA, Klein T, et al. Joint Clinical Practice Guideline on Benzodiazepine Tapering: Considerations When Risks Outweigh Benefits. J Gen Intern Med. 2025. PMID 40526204.
5. Croskerry P. The importance of cognitive errors in diagnosis and strategies to minimize them. Acad Med. 2003. PMID 12915363.
6. Darker CD, Sweeney BP, Barry JM, et al. Psychosocial interventions for benzodiazepine harmful use, abuse or dependence. Cochrane Database Syst Rev. 2015. PMID 26106751.
7. FDA. FDA requiring Boxed Warning updated to improve safe use of benzodiazepine drug class. 2020. fda.gov.
8. Kelly JF, Westerhoff CM. Does it matter how we refer to individuals with substance-related conditions? A randomized study of two commonly used terms. Int J Drug Policy. 2010. PMID 20005692.
9. Lader M. Benzodiazepine harm: how can it be reduced? Br J Clin Pharmacol. 2014. PMID 22882333.
10. Parr JM, Kavanagh DJ, Cahill L, et al. Effectiveness of current treatment approaches for benzodiazepine discontinuation: a meta-analysis. Addiction. 2009. PMID 18983627.
11. Rochon PA, Gurwitz JH. Optimising drug treatment for elderly people: the prescribing cascade. BMJ. 1997. PMID 9366745.
12. Soyka M. Treatment of Benzodiazepine Dependence. N Engl J Med. 2017. PMID 28328330.
13. Tannenbaum C, Martin P, Tamblyn R, et al. Reduction of inappropriate benzodiazepine prescriptions among older adults through direct patient education. JAMA Intern Med. 2014. PMID 24733354.
14. Voshaar RC, Couvée JE, van Balkom AJ, et al. Strategies for discontinuing long-term benzodiazepine use: meta-analysis. Br J Psychiatry. 2006. PMID 16946355.