Key points are not available for this paper at this time.
OBJECTIVES: To assess the clinical and cost-effectiveness of zaleplon, zolpidem and zopiclone (Z-drugs) compared with benzodiazepines. DATA SOURCES: Electronic databases, reference lists of retrieved articles and pharmaceutical company submissions. REVIEW METHODS: Randomised controlled trials (RCTs) that compared either benzodiazepines to the Z-drugs or any two of the non-benzodiazepine drugs in patients with insomnia were included in the review. Data on the following outcome measures were considered: sleep onset latency, total sleep duration, number of awakenings, quality of sleep, adverse effects and rebound insomnia. A search was also undertaken for any study designs that evaluated issues related to adverse events (e.g. dependency and withdrawal symptoms). Full economic evaluations that compared two or more options and considered both costs and consequences including cost-effectiveness, cost-utility analysis or cost-benefit analysis undertaken in the context of high-quality RCTs were considered for inclusion in the review. RESULTS: Twenty-four studies, involving a total study population of 3909 patients, met the inclusion criteria. These included 17 studies comparing a Z-drug with a benzodiazepine and seven comparing a Z-drug with another Z-drug. The diversity of possible comparisons and the range of outcome measures in the review may be confusing. Outcomes were rarely standardised and, even when reported, differed in interpretation. In addition, variations in assessment and variety in the level of information provided make study comparisons difficult. As a result, meta-analysis has been possible on only a small number of outcomes. However, some broad conclusions might be reached based on the limited data provided. The existing published economic literature in this area is very limited. No relevant economic evaluations were identified for inclusion in the review. The industry submissions did not include detailed evidence of cost-effectiveness. Given the lack of robust clinical evidence, no economic model describing the costs and benefits of the newer hypnotic drugs for insomnia was developed. The systematic review provided in this report suggests that an agnostic approach to cost-effectiveness is required at this stage. In the short-term, no systematic evidence is available concerning significant outcome variations between either the different classes of drugs or between individual drugs within each class. Within this short-term horizon, the one element that does vary significantly is the acquisition cost of the individual drugs. CONCLUSIONS: The short-acting drugs seem equally effective and safe with minor differences that may lead a prescriber to favour one over another in different patients. There is no evidence that one is more cost-effective than any other. Analysis of the additional costs to the NHS, depending on the rate of change from benzodiazepine prescriptions to Z-drug prescriptions, at current levels of hypnotic prescribing, range from GBP2 million to GBP17 million per year. There are clear research needs in this area; in particular, none of the existing trials adequately compare these medications. It is suggested that further consideration should be given to a formal trial to allow head-to-head comparison of some of the key drugs in a double-blind RCT lasting at least 2 weeks, and of sufficient size to draw reasonable conclusions. We would also recommend that any such trial should include a placebo arm. It should also collect good-quality data around sleep outcomes and in particular quality of life and daytime drowsiness. We do not believe that any formal study of risk of dependency is feasible at present. Finally, the management of long-term insomnia is suggested for further investigation: considering the frequency of this symptom and its recurring course, the short-term trial of medication and lack of long-term follow-up undermine attempts to develop evidence-based guidelines for the use of hypnotics in this condition, or indeed for its whole management.
Dündar et al. (Tue,) studied this question.