374 in these claims with data on the general use of the benzodiaze- pines in Sweden. No medical records were available for analysis. Compared with the general population using benzodiaze- pines, those making claims (table) differed in: the high use of anxiolytic compared with hypnotic benzodiazepines, both sexes were younger, and the female preponderance was less. Except for the high number of lorazepam claims, the drug pattern agreed with use during the past decade. Among the stated adverse reactions, withdrawal, dependence, and anxiety predominated (67%). Our findings suggest that the alleged problems with benzo- diazepines may be due to the underlying reasons for which these drugs were prescribed, rather than being a general benzodiazepine phenomenon (all demands were rejected by the insurance scheme). In household surveys of anti-anxiety sedative drug use in European countries, the largest decrease was found in the UK in 1990 (45% use) compared with 1981 (11’2%).3’’ With the heated debate on the use of benzodiaze- pines, for ethical reasons hard demands must be placed on the authorities to have surveillance systems installed that will guarantee patients access to medically-justified benzodiaze- pines, and restrict unauthorised (illicit) use. It may be questioned if this balance is right in Sweden and elsewhere today. The 1990 survey pointed at undertreatment of people who fulfilled criteria for the treatment of anxiety disorders.4 Ulf Bergman, Mårten Myrhed Department of Clinical Pharmacology, Karolinska Institutet, Huddinge University Hospital, S-141 86 Huddinge, Sweden 1 Bergman U. Epidemiological aspects on the use and abuse of benzodiazepines. In: Hearing on benzodiazepines: benefits and risks. Stockholm, December 1991, National Board of Health and Welfare & Medical Products Agency, pp 5-15 (in Swedish). 2 Nordic Statistics on Medicine 1987-1989. NLN publication no 30. Uppsala: Nordic Council on Medicines, 1990. 3 Balter MB, Manheimer DI, Mellinger GD, Uhlenhuth EH. A cross- national comparison of anti-anxiety sedative drug use. Curr Med Res Opin 1984; 8 (suppl 4): 5-18. 4 Balter MB. Benzodiazepine Use: 1990 survey data and benefit risk issues: results from a national epidemiological study. 5th World Congress of Biological Psychiatry; Florence, 1991. Multiple Sclerosis Society’s advertisement SiR-I do not think that I am alone in objecting to the Multiple Sclerosis Society appeal on the back cover of The Lancet of July 3. Phrases such as "a hope in hell" and "incurable" may have their place in modern advertising practice, but they contribute to the popular misconception that multiple sclerosis is a uniformly disabling disease with a uniformly terrible prog- nosis. The society offers much to a newly diagnosed patient, not least through its research commitment; but the fear engendered by such publicity may counteract its supportive role. Antony Johansen Whipps Cross Hospital, London E11 1NR, UK MS Society’s reply SiR-All advertising used by the Multiple Sclerosis Society is first thoroughly researched, and this research includes people with multiple sclerosis. With the present campaign this research showed very clearly that it was likely to be well received by the general public and the subsequent results have shown this to be so. Our advertising campaigns are monitored throughout their lifetime and should this monitoring reveal that the campaign is communicating the wrong message action would be taken. The MS Society is the major funding body for MS research in the UK and at the same time is being required to commit far higher levels of funding to welfare and support services. We rely entirely upon voluntary donations, so a high public profile is important. Our advertising and other pro- motional activities have achieved this objective. John Walford, General Secretary Multiple Sclerosis Society of Great Britain and Northern Ireland, 25 Effie Road, London SW6 1EE, UK Should prescribers have a say? SIR—To achieve international harmonisation of data sheets a working group of the Council for International Organisations of Medical Sciences (CIOMS) was established (CIOMS III). The working group consists of roughly equal numbers of persons from the pharmaceutical industry and the regulatory authorities. However, although there are physicians in the group there are no representatives from the prescribers. The absence of prescriber representatives is strange since they are the customers who use the information contained in the data sheets. The findings of CIOMS III will not be binding on countries;’ but if they follow the precedence established with CIOMS I and II they will be accepted by most developed countries. M D B Stephens 49 King’s Court, Bishop’s Stortford, Hertfordshire CM23 2AB, UK 1 International reporting of adverse drug reactions: find report of CIOMS working group. Geneva: CIOMS, 1990: 5. Tolerance and withdrawal with zolpidem SIR—Zolpidem is a non-benzodiazepine short-acting hyp- notic. Clinical trials and animal pharmacology mostly suggest poor liability to tolerance and related events,1,2 but baboon data show no difference in tolerance and withdrawal compared with benzodiazepines.3 We have observed two cases of tolerance to zolpidem’s hypnotic effects with withdrawal phenomena. Case 1 (F/60, histrionic and paranoid personality disorder treated with zolpidem 10 mg at night, admitted as emergency for supposed zolpidem overdose). She was poorly oriented in time and space, dysarthric, confused, grossly disinhibited, and showed unsteady gait. After an agitated night (promazine 50 mg was administered for sedation), she was cooperative and oriented but anxious and restless, with anterograde amnesia. Limb, hand, and perioral tremor, and muscle twitching and myoclonic jerks could be seen, and she also reported diplopia, abdominal, and gastric pain with swallowing difficulties. 18 h after drug ingestion a generalised self-limited seizure occurred. Intravenous infusion of chlordesmethyldiazepam 5 mg and flunitrazepam 20 mg orally at night resolved the physical and neurological symptoms and the patient could sleep. She was discharged 3 days later on clonazepam 2 mg daily. Residual symptoms were muscular and abdominal pain. In the last 2 months she had increased the zolpidem dose up to 100 mg (ten tablets) because of tolerance to the hypnotic effect with wakening after 2-3 h. Abstinence phenomena (anxiety, tremor, sweating, nausea, difficulties in swallowing, gastric and ab- dominal pain) emerged during the day, but the patient attributed them to lack of a "good" sleep. A similar episode, with withdrawal seizures, had occurred last year, 18-20 h after an overdose of triazolam, which the patient had abused. Other causes of seizures had been excluded. Case 2 (F/31, previous diagnosis of major depressive episode treated for residual insomnia with zolpidem 20 mg daily). After about 2 months, tolerance to the hypnotic effect, with wakening after 3 h occurred. This induced the patient to increase doses up to 40-50 mg. After a few months, beside tolerance to the