Analytical approach:
This study used an economic model to estimate the cost-utility outcomes for eszopiclone treatment and the main data source was a clinical trial. The time horizon was six months and the authors stated that the perspective was that of society.
Effectiveness data:
The effectiveness data were from a single clinical trial that lasted for six months. The patients had chronic primary insomnia that was diagnosed using the Diagnostic and Statistical Manual of mental disorders (DSM) IV and other sleep criteria. The Insomnia Severity Index (ISI) was used to assess the severity and impact of insomnia symptoms on patients, and the Work Limitations Questionnaire (WLQ) was used to measure the percentage of productivity lost by the patient and the patient’s ability to perform their job.
Monetary benefit and utility valuations:
In the clinical trial, the Short Form (SF) 36 Health Survey was used to measure health-related quality of life. A transformation method, from a published study, was used to transform the SF-12 scores, a subset of the SF-36, into European Quality of life (EQ-5D) questionnaire scores. This published study mapped the scores of 13,000 US patients in the National Medical Expenditure Survey between the SF-12 and the EQ-5D.
Measure of benefit:
The measure of benefit was the number of quality-adjusted life-years (QALYs).
Cost data:
The cost categories were the direct medical costs and productivity losses, which included absenteeism and presenteeism, defined as lost productivity while at work. The average wholesale price of eszopiclone was reduced by 18.3%, to reflect the over-estimation of actual pharmacy acquisition costs, and a monthly dispensing fee was added, to reflect actual practice, for the drug costs. The WLQ was used to calculate the monetary value of lost productivity. The medical resource use was collected during the clinical trial, using the Health Utilisation Questionnaire. To estimate the physician visits in a non-experimental setting, participants were categorised according to their sustained remission from insomnia and those with remission were assumed to have fewer visits than those without; the reduced visits and their costs were based on a retrospective claims database. Absenteeism was assessed in the same way, using the MarketScan Health and Productivity Management Database. The direct costs not covered by a third-party payer, prescription co-payments, and the impact of insomnia on family members and caregivers were not included. The costs were adjusted to 2006 US dollars ($).
Analysis of uncertainty:
Univariate and multivariate sensitivity analyses were conducted. Bootstrapping of the clinical trial data was performed 5,000 times. A scenario analysis was performed to investigate the impact of the exclusion of out-patient or productivity costs or both, on the results.