23rd March 2012
An observational study has suggested that patients who received hypnotic prescriptions had a significantly increased risk of dying over an average follow-up of 2.5 years, compared to those who didn’t receive hypnotic prescriptions. A dose-response relationship was seen: the risk was increased around 3.6 times in patients who were prescribed fewer than 18 doses of hypnotic per year, and around 5.3 times in those prescribed more than 132 doses/year.
Level of evidence:
Level 2 (limited quality patient-oriented evidence) according to the SORT criteria.
Prescribers and prescribing managers should ensure that hypnotics are only prescribed when absolutely necessary, in accordance with national guidance. Non-pharmacological interventions should be considered first-line, for example, advice on appropriate routines to encourage good sleep; avoiding stimulants; maintaining regular sleeping hours with a suitable environment for sleep; and cognitive behavioural therapy. (See NICE TA77.)
The Committee on Safety of Medicines advises that benzodiazepine hypnotics should be used only if insomnia is severe, disabling or causing the patient extreme distress. The lowest dose that controls symptoms should be used, for a maximum of four weeks and intermittently if possible. NICE guidance on zaleplon, zolpidem and zopiclone (the so called ‘Z drugs’) also recommends that when, after due consideration of the use of non-pharmacological measures, hypnotic drug therapy is considered appropriate for the management of severe insomnia interfering with normal daily life, hypnotics should be prescribed for short periods of time only, in strict accordance with their licensed indications. Patients who have been taking hypnotics long-term should be reviewed with the aim of reducing and stopping treatment.
Various approaches to reducing hypnotic prescribing can achieve significant success: see ‘Moving towards personalising medicines management: Improving outcomes for people through the safe and effective use of medicines’.
What is the background to this?
According to the authors of the study, at least 24 published studies have examined the association between hypnotic use and mortality, with the majority concluding that using hypnotics increased mortality. In addition, several studies have shown an association between hypnotics and cancer deaths. Studies have generally not reported which specific hypnotics were used, or included sufficient data on newer short-acting benzodiazepines such as zolpidem and zaleplon.
This large US matched cohort study investigated the association between specific commonly used hypnotics and mortality or major cancer (excluding non-melanoma skin cancer). 10,529 patients who received at least one hypnotic prescription were matched with 23,676 controls who did not any receive hypnotic prescriptions. Records were examined to see if patients subsequently died or were diagnosed with major cancer.
What does this study claim?
The study found that patients prescribed hypnotics had a significantly increased risk of dying compared with those who were not prescribed hypnotics. Over a mean follow-up of 2.5 years, 6.1% of hypnotic users died, compared with 1.2% of non-users. A dose-response relationship was seen, with an increased risk of death seen even in patients using fewer than 18 hypnotic doses per year. For patients prescribed 0.4 to 18, 18 to 132 and more than 132 doses/year, hazard ratios (HRs) were 3.60 (95% confidence interval [CI] 2.92 to 4.44), 4.43 (95% CI 3.67 to 5.36) and 5.32 (95% CI 4.50 to 6.30) respectively.
Eight commonly used hypnotics (zolpidem, temazepam, eszopiclone, zaleplon, triazolam, flurazepam, barbiturates and sedative antihistamines) were examined. Patients prescribed more doses of hypnotic per year were more likely to develop a new major cancer. In those taking more than 132 doses/year the hazard ratio was 1.35 (95% CI 1.18 to 1.55) (see table 1 ). When the hypnotic agents were analysed individually, almost all were found to be associated with an increased risk of death (see table 2 ).
While this study cannot prove conclusively that hypnotics increase the risk of mortality or cancer, it does raise concerns, particularly as the risk of death was raised even in patients who used fewer than 18 doses/ year. The study adds to the existing evidence about the harms of hypnotics and provides another reason to exercise caution when considering prescribing hypnotics, even when used according to national guidance (at low doses and for short periods of time). It also highlights the importance of reviewing patients who are already taking hypnotics, with a view to stopping treatment.
Risks associated with hypnotic drugs have been well recognised for many years. As long ago as 1988, the Committee on Safety of Medicines advised that benzodiazepine hypnotics should be used only if insomnia is severe, disabling or causing the patient extreme distress. The lowest dose that controls symptoms should be used, for a maximum of four weeks and intermittently if possible. NICE guidance on zaleplon, zolpidem and zopiclone makes similar recommendations. More recently, the MHRA highlighted again the issues regarding addiction to benzodiazepines in the July 2011 edition of Drug Safety Update, which points to a structured review by the National Addiction Centre of available evidence on prescribing trends in benzodiazepines and ‘Z drugs’ and current evidence for the extent of dependence and harms attributable to these drugs. Despite these national safety warnings and guidance, overall prescribing of hypnotics is not decreasing (see Figure).
Figure: Trends in prescribing of hypnotics in general practice in England (July 2006 – Sept 2011)
A single observational study, such as this one, can prove only association not causation and observational studies are prone to confounding. Unlike in the setting of a randomised controlled trial, in ‘real life’, treatment plans are chosen, changed, or actively not chosen in the light of individual patients’ risk factors, preferences and tolerability or response to other drugs. Thus observed differences in outcomes may be due to differences among the patients, not only the different treatments.
In this study the authors adjusted for age, gender, smoking, body mass index, ethnicity, marital status, alcohol use and prior cancer. However, residual confounding cannot be ruled out. Although multiple strategies were used to address confounding by health status, it was not possible to control for depression, anxiety and other mental health that may increase the risk of death through pathways independent of the hypnotics.
Hypnotics are included in the QIPP document: ‘Key therapeutic topics – Medicines management options for local implementation’. A prescribing comparator and a set of e-learning resources are also available to support implementation of this topic. More information on insomnia can be found within the CNS and mental health NPC e-learning materials and on NHS Evidence.
Design: One-to-two matched cohort study.
Patients: A large US database was used to identify patients who had at least one prescription for a hypnotic for insomnia who survived for three months or more. Patents diagnosed with major cancer (excluding non-melanoma skin cancer) were excluded.10,529 patients who received hypnotic prescriptions were matched with 23,676 controls with no hypnotic prescriptions, and followed for an average of 2.5 years between January 2002 and January 2007. The mean age of patients was 54 years.
Intervention and comparison: Records were examined to see if patients subsequently died or were diagnosed with major cancer. Data were adjusted for age, gender, smoking, body mass index, ethnicity, marital status, alcohol use and prior cancer. Multiple strategies were used to address confounding by health status.
Outcomes and results: Patients prescribed any hypnotic had significantly elevated hazards of dying compared to those prescribed no hypnotics. A dose-response relationship was seen. Hypnotics were also associated with a significantly increased risk of cancer, except in patients taking fewer than 18 doses/year.
|Any hypnotic: doses/year||Deaths||Cancers|
|P value||HR (95% CI)||P value||HR (95% CI)|
|0.4 to 18Mean 8
|<0.001||3.60(2.92 to 4.44)||0.086Not significant||0.86
(0.72 to 1.02)
|18 to 132Mean 57
|<0.001||4.43(3.67 to 5.36)||0.022||1.20(1.03 to 1.40)|
|<0.001||5.32(4.50 to 6.30)||<0.001||1.35(1.18 to 1.55)|
Hazard ratios were elevated in separate analyses for several common hypnotics, including zolpidem, temazepam, eszopiclone, zaleplon, other benzodiazepines, barbiturates and sedative antihistamines.
|Hypnotic||No. patients||No. deaths||HR (95% CI)|
|4.82 (4.06 to 5.74)|
|4.98 (4.05 to 6.14)|
|30.62 (12.90 to 72.72)|
|3.75 (2.29 to 6.12)|
|4.50 (1.83 to 11.10)|
|2.21 (0.98 to 4.98)Not significant|
|2.78 (1.57 to 4.92)|
|4.57 (3.01 to 6.94)|
Results were robust within groups suffering each comorbidity, indicating that the death and cancer hazards associated with hypnotic drugs were not attributable to pre-existing disease.
Sponsorship: Not commercially funded
Please comment on this rapid review using our feedback form.
Make sure you are signed up to NPC Email updates — the free email alerting system that keeps you up to date with the NPC news and outputs relevant to you.