Community survey of long term daytime use of benzodiazepines

BMJ 1994; 309 doi: (Published 02 July 1994) Cite this as: BMJ 1994;309:27
  1. N Wright,
  2. R Caplan,
  3. S Payne
  1. Peter Hodgkinson Centre, Lincoln County Hospital, Lincoln LN2 5QY. Mapperley Hospital, Nottingham NG3 6AA.
  1. Correspondence to: Dr N Wright, Department of Psychiatry for the Elderly, Leicester General Hospital, Leicester LE5 4PW.
  • Accepted 22 February 1994

Media coverage suggests that many people are addicted to benzodiazepines and keen to stop taking them. Our recent efforts to recruit subjects into a programme offering withdrawal, however, had a disappointing response. As a result we initiated this survey. Our decision to study patients taking benzodiazepines long term reflects the current medical opinion that long term use is likely to cause problems.1

Patients, methods, and results

Subjects were recruited from a general practice serving 13 000 patients. The catchment was mixed, including prosperous suburbs and council estates. Using the practice's computerised records for repeat prescriptions we identified all patients taking benzodiazepines during the day and included those in the survey who had been prescribed the drugs for more than one year. Patients were interviewed at home. Psychiatric history and detailed histories of benzodiazepine use were taken. Patients completed the benzodiazepine withdrawal symptom questionnaire.2 Data were analysed using X2 and two sample t tests.

Sixty five patients had been prescribed benzodiazepines for more than one year (point prevalence 0.5%). Eight refused to be interviewed and two were not contactable. Fifty five were therefore interviewed (14 men, 41 women; mean age 57.4 years). Demographic details of those not interviewed were not significantly different from those of the study cohort.

The only variable differentiating between patients keen to stop their treatment and the rest of the cohort was age: those desiring to stop were younger (53.8 years v 61.4 years (95% confidence interval of the difference 1.4 to 13.8 years), P=0.037). Patients who had suffered withdrawal symptoms (physical dependence) as measured by the questionnaire were less likely to think that they still benefited from the benzodiazepine (11/17 v 32/35, X2=5.72, P<0.02). They were younger than those with no history of physical dependence (50.9 years v 60.3 years (1.8 to 17.1 years) P=0.017) but were not distinguishable by any other variable. Underusers were significantly older than the rest of the study cohort (64.2 years v 55.1 years (2.8 to 15.4 years), P=0.0074).

Patterns of use, attitudes towards use, and problems with dependence among patients taking benzodiazepines long term during day

View this table:


The small number of patients identified in this survey confirms the continued decline in the prescription of benzodiazepines for anxiolysis.3,4 Twenty four patients reported having taken benzodiazepines long term on previous occasions. This should encourage optimism about the prospects of successfully stopping treatment, although it also suggests that patients may start taking benzodiazepines again. It is often forgotten that only a minority of patients taking benzodiazepines long term experience withdrawal symptoms (33% in this study).5 A lack of withdrawal symptoms is likely to have enable patients to stop taking the drugs. Continued benefit and a lack of desire to stop treatment probably both work to ensure continued long term use for many patients.

Only 15 patients (27%) had not come into contact with psychiatric services before. Concern that services could be overwhelmed if all patients taking benzodiazepines long term were referred to psychiatric teams seems to be unfounded.

Our findings fail to confirm popular preconceptions about long term use of benzodiazepines. We found a small number of long term users. Some surprising attitudes were expressed, in particular the desire of almost half of the group to continue taking their drugs. There ware reassuring findings about patterns of use, continued benefit, and frequency of withdrawal symptoms. The association of a desire to stop treatment and experience of withdrawal symptoms with younger age suggests that greater caution should be exercised when considering benzodiazepines for daytime use in younger patients. If community based research is to establish the most appropriate use of benzodiazepines large catchment populations will be needed for sufficient numbers to be identified, particularly for analysis of subgroups.

Our findings do not conform to the benzodiazepine stereotype that is often presented in the media, but they do suggest reasons for the poor response to our withdrawal programme. We suggest that media coverage of this topic should be more restrained.

We thank the general practitioners at the Birchwood Health Centre and Paul Novak.


  1. 1.
  2. 2.
  3. 3.
  4. 4.
  5. 5.
View Abstract