Skip to main menu | Skip to content

You are in the Publications section

How to recognise and treat benzodiazepine dependence

A guide for General Practitioners

Dr Andrew Byrne. Originally published in the Medical Observer (Australia), July 25, 1997



GPs encounter two distinct types of tranquilliser dependence. These have been termed therapeutic and 'morbid'. Most of the drugs come from prescribed sources and both types of presentation benefit from intervention at the appropriate time.


After brief clinical trials at the University of Texas in 1959, chlordiazepoxide (Librium) was the first of the benzodiazepines to be marketed for insomnia and anxiety. It took less than a decade for this class of drugs to replace over 90% of a market previously dominated by the barbiturates. Another decade was to pass, however, before it became generally accepted that the newer drugs had many problems of their own, greatly limiting their usefulness.

Although possessing a much lower primary toxicity than barbiturates, it is now known that tolerance, dependence and withdrawals are extremely common in those taking benzodiazepines.

The current prescribing information is candid and would deter many new patients. In a page of the smallest discernible print, we are told that these drugs are no longer recommended for chronic anxiety or insomnia. Benzodiazepines induce abnormal sleep patterns and cannot restore sleep pathology on their own. With few patients fitting current diagnostic criteria, some authorities would limit the use of these drugs to panic attacks, alcohol withdrawal and status epilepticus. But practical alternatives are scarce.

Yet we are dealing with a demanding public who still request these drugs for a variety of reasons. When confronted with a patient requesting benzodiazepines, we need to look both at the individual clinical setting as well as the bigger picture.

How can I help this patient? Is it possible to establish a therapeutic relationship without just prescribing 'on demand'? Where does 'harm reduction' fit into all this?

Pill abusers are more often female while opiate users are more likely to be male. GPs are all familiar with the extremes: 'morbid' abusers are usually younger patients who may be using other drugs as well. They often take far higher than therapeutic doses. At the other extreme is an older patient who is taking therapeutic or slightly higher doses of benzodiazepines with or without alcohol. There is also a grey area in between. We now know that patients in both groups are in danger of serious complications and hence may need intervention.

Are they dependent? What are the risks? Can we intervene in any useful way to reduce those risks?


Few GPs would insist that elderly folk on modest doses of hypnotics should immediately withdraw their medication. However, with new evidence of an association between low dose benzodiazepine use and serious falls, older patients must be allowed to reduce their dose where this is possible. Liquid filled capsules ('footballs') should generally be avoided as they cannot be bisected. Better agents to use are those available in scored tablets and graduated lower strengths (eg oxazepam 15mg or diazepam 2mg). Some believe that longer acting forms such as diazepam are less prone to abuse than the short half-life drugs.

It has been shown in controlled trials that 'brief interventions' by GPs are effective at six and twelve months follow-up. This is true for smoking and alcohol consumption and is likely to be so with benzodiazepine use and abuse. But how do we 'intervene'?


For young people taking large quantities of drugs, it may be impossible to take a clear history. Some of these patients wander about the point, they make up stories and ramble on. They may address staff by first names and they also may move surgery furniture around. Where patients are frankly intoxicated, we should not endeavour to do the impossible, but take steps to deflect attention away from the prescription pad.

We should take the blood pressure and at the same time look for venous scarring. There is a polite, caring way of doing this. "I see we have had a little problem in the past." "Have you ever had a thrombosis?" "Would you like me to check your other veins?" A brief, informal test of memory will often reveal problems. Tolerance develops to all the effects of benzodiazepines apart from the amnesia, which is dose-dependent. This may be associated with such interesting phenomena as flash backs and the 'invisible man' syndrome.

One approach might be to say: "I believe that you may have a major problem and I would like to help you with it. This may take some time, and I would like you to sit in the waiting room for a short while. Would you like a cup of tea?" A less charitable approach might be to hand them a 'bottle' and ask for a urine test.

If patients are in withdrawal and they have been waiting for some time already, refusal can be met with anger and even violence. There may be warning signs of this with pacing, wringing of hands, jaw clenching, etcetera. In these exceptional cases, one might prescribe a minimal quantity of the patient's drug of choice such as 25 x 15mg oxazepam tablets. Advice should be given to return when they run out to 'talk about their problem'. The patient who is not yet amenable to intervention will usually leave the premises promptly. They know that you have recognised their problem and have been positive about treatment. They also know that you do not write prescriptions on demand. This is more likely to instil respect in such patients. 'Easy-touch doctors' are never doing their patients a favour. They encourage 'doctor shopping', an issue which is belatedly being addressed by state authorities.


So when our patient tells us frankly about their ten Normison, two Rohypnol, five Xanax and one bottle of 'reserve sherry' per day what to we do? Having gained their confidence, we need to delineate the history of drug use and how they feel about it. Some are only abusing tranquillisers. Other also abuse opiates, cocaine or other stimulants. Is the drug use compulsive? Is it actually harmful? Are they still in control? Do they want to stop altogether? If so, over what period? Have they experienced detoxification before?

Unlike methadone or nicotine patch treatment, we cannot recommend any single modality to these patients since there is still no proven management for benzodiazepine dependence. It is easier to determine what not to do. To repeat the 'say no to drugs' philosophy is trite, negative and unhelpful. Many pill users, however, can and docurtail their habit in a matter of days or weeks with few serious withdrawal symptoms.


In common with all other addictions, admitting the problem is the first step. Deciding when and what to do about it comes next. Discussing these issues is well suited to the GP consultation, in the same way as smoking cessation or drinking problems.

One of the doctor's responsibilities is to give the patient information on which to base their personal decisions. Some patients are ignorant of the most fundamental properties of the drugs they are taking. By education we can dispel myths and misconceptions, thereby assisting the patient to make informed decisions about how to use (or not use) such drugs.

Many 'benzo' addicts are unduly concerned about fitting if they cut down their dose. After years of using pills in a binge manner, if a patient has never had a fit, then they are at low risk of convulsions. On the other hand, those who do take fits need special attention to graduated dose reductions. Some impostors use this as a ruse to obtain drugs from sympathetic or naive doctors.

Another common misapprehension is that you 'cannot die from benzodiazepines'. This is not the case, and, while pure sedative overdose is uncommon, deaths involving benzodiazepine combinations are tragically frequent in Australia today. They cause respiratory depression which can be profound, especially in combination with opioid analgesics. Although intravenous flumazenil is an effective agent for benzodiazepine reversal, most deaths unfortunately occur before medical intervention is possible.


It is clearly unwise to give a patient who is out of control with drug use a retail quantity of drug. If immediate detoxification is not possible then smaller quantities on a regular basis may be more appropriate. This has been used in out-patient reduction regimens, usually with diazepam. The harm reduction philosophy allows those who are addicted access to limited quantities of safe drugs within a medical framework.

Pharmacists are usually prepared to participate in more frequent dispensing of smaller quantities. Some unstable patients benefit from daily supervised dosing, while others can take a week's supply at a time. The patient should have some input into the goal setting which should not be imposed arbitrarily by the doctor. Some limits and objectives should be agreed. These may include an attendance diary, supervised urine tests, family and work responsibilities or an arranged interview or follow-up with a D&A specialist.

Initial doses should be according to the patient's current tolerance. In those requiring high doses (over 30mg diazepam daily or equivalent), this may be confirmed by an examination 3 hours after witnessed dosing. Reductions should be no more than 25% of the daily dose at 2 week intervals. The longer and slower the reductions, the more likely the process is to succeed long term. After years of pill taking, another month or even two on treatment may not be excessive.


One setting in which clear guidance is available is for those who are already on methadone. Maintenance treatment is the most acceptable and effective option for compulsive heroin users. Although the majority cease the use of other drugs after sufficient time in treatment, some continue to 'doctor-shop' and self medicate. Before prescribing any formal withdrawal regimen, it is essential to ascertain whether the patient already has a narcotic authority, and who is their doctor. With a current authority issued, it would be unwise and possibly illegal to prescribe another controlled drug without permission.

The patient's methadone treatment should be reviewed in conjunction with their prescriber who may be a GP, psychiatrist or a clinic doctor. 'Shared care' has been used successfully for drug addicts overseas. The review may include an examination of dose levels, attendance record and if necessary a trough methadone level which is now a simple, routine test. Pupillary size is very often inversely proportional to narcotic levels.


The writing is on the wall for benzodiazepine use. Greater numbers of drug deaths, increased risk behaviour for HIV and other communicable diseases, falls and accidents in the elderly are all previously unsuspected hazards. Recent years have seen a small but sustained reduction in the use of these drugs. Authorities have already legislated in certain areas where the profession was tardy (clonazepam authorities, New York triplicate prescription system, de-listing of flunitrazepam, etc). These moves all greatly reduced the consumption of tranquillisers, albeit at the expense of some professional freedom.

Only further experience will tell us whether the newer agents for anxiety and depression (SSRIs etc.) are safer and more effective than benzodiazepines. Meantime, non-chemical means have to be relied upon in most cases as doctors and patients await the perfect anxiolytic.


Trickett, Shirley. Coming off tranquillisers and sleeping pills . Thorsons, UK. Revised edition 1991.

Faust, Beatrice. Benzo Junkie. Penguin, Australia. 1993 Rational Use of Benzodiazepines. Policy Position Statement,Australian Professional Society on Alcohol and other Drugs. Ed. Mant A.

Psychotropic Drug Guidelines, 3rd Edition, July 1995. VMPF Therapeutic Committee on behalf of the Victorian Drug UsageAdvisory Committee. Cumming R, Klineberg R. Psychotropics, thiazide diuretics and hip fractures in the elderly.1993 158:414-17.

Byrne A. Benzodiazepines: the end of a dream. Aust Fam Physician 1994 Aug; 23(8):1584-1585

While every care has been taken to ensure that the content of this page is accurate, the author and publisher do not accept legal liability for any problems arising from the implementation of the various treatment strategies outlined therein.
This work is copyright. Apart from any fair dealing for the purposes of private study, research, criticism or review, as permitted under copyright law, no part may be reproduced by any process without written permission from the publisher.
© Andrew James Byrne 1997
© andrew byrne     +61 2 9319 5524
Valid XHTML 1.0 Transitional Mostly CSS validated but hacked for IE!