Analytical approach:
A decision analytic model was used to compare the costs and outcomes for the different sedatives. The time horizon was 28 days. The model assumed that, if a patient tolerated the sedative, they survived or died based on the model inputs. However, if the patient experienced sedation-related adverse effects they would cross over to the comparator sedative. The authors stated that the study perspective was that of the hospital.
Effectiveness data:
The majority of the effectiveness estimates were derived from two published, randomised controlled trials (RCTs). Some supplementation with data from other published studies was necessary. The authors did not report any search methods or inclusion criteria. The main clinical parameters were hospital mortality, medication failure leading to crossover, ventilator days, and intensive care unit length of stay.
Monetary benefit and utility valuations:
Not relevant.
Measure of benefit:
The two measures of benefit derived were mechanical ventilator-free days (for all patients) and mechanical ventilator-free survival (days for survivors only).
Cost data:
The cost categories included medications, tests, intensive care and hospital ward stays, and physician costs. The cost data were obtained from a variety of sources including national databases, institutional costs, and published sources. Most of the resource use data were derived from the RCTs which provided the effectiveness estimates. All costs were adjusted to year 2007 US dollars ($) using the medical care component of the US Consumer Price Index. The cost estimates were subject to extensive sensitivity analysis.
Analysis of uncertainty:
The authors performed one-way and two-way sensitivity analyses on the model parameters using ranges from published studies or using plus and minus 50% of the base-case estimates. Also, the main results were presented for probabilistic analyses, generated using Monte Carlo simulation. The details of the distributions used in these probabilistic analyses were presented.