Analytical approach:
The analysis was based on a decision-tree model, with a short time horizon of seven days, reflecting the hospital stay. The authors stated that the perspective of the hospital was adopted.
Effectiveness data:
A literature review was undertaken to identify head-to-head randomised controlled trials (RCTs) of the two drugs. Three trials were identified and the weighted averages of their clinical evidence were calculated. Discontinuation rates were only available from the largest trial. The key details of these trials were reported. The main input was the probability of seizure or adverse reaction with the two drugs.
Monetary benefit and utility valuations:
Not considered.
Measure of benefit:
The summary benefit measure was a successful seizure prophylaxis regimen. A successful regimen was one on which a patient did not have a seizure nor an adverse drug reaction that resulted in a change in therapy.
Cost data:
The economic analysis included the costs of levetiracetam, phenytoin, carbamazepine, the stay in the intensive care unit, and phenytoin laboratory monitoring. The resource consumption was the expert opinion of two intensive care pharmacists at the authors’ institution or from published literature. The drug costs were their average wholesale prices for the regimens, while other costs were from the Veterans Affairs Decision Support Services (DSS) database. The price year was 2008 and all costs were in US dollars ($).
Analysis of uncertainty:
One-way sensitivity analyses were carried out on all the model inputs, using published and assumed ranges of values. A second-order Monte Carlo simulation was performed, using 10,000 iterations and predetermined probability distributions for the types of input; beta for probabilities, gamma for costs, and normal for resource use. Cost-effectiveness acceptability curves were generated.