Analytical approach:
A decision tree was used to synthesise the data from published clinical trials for the four treatment regimens. For each treatment arm of the decision tree, a patient followed a treatment path either with or without complications. The authors stated that the perspective was that of a health care provider.
Effectiveness data:
The clinical evidence was identified from recent available anticoagulation studies that compared the four strategies. The procedural techniques, complication rates, and treatment regimens were closely based on those of the Superior Yield of the New strategy of Enoxaparin, Revascularization and Glycoprotein IIb/IIIa inhibitors (SYNERGY), the Fifth Organization to Assess Strategies in Acute Ischemic Syndromes (OASIS-5), and ACUITY trials, which were the main motivation and data sources for this study. The outcomes and probabilities associated with each treatment strategy were derived from point estimates reported in these three trials. A literature review was conducted to obtain the data on consistent anticoagulation in patients.
Monetary benefit and utility valuations:
Not relevant.
Measure of benefit:
The measure of benefit was patients treated without complications, as this was the only clinically significant difference between the treatments.
Cost data:
The cost categories were the treatment, drug acquisition, and complications of the interventions. The drug acquisition costs were from the wholesaler-purchasing database of the Virginia Commonwealth University Medical Center. The major complication costs were based on diagnosis-related group data. Physician fees were estimated by assigning a Current Procedural Terminology (CPT) code to all complications. The minor complication (bleeding) costs were from hospital blood bank reports. The resource use data were from published clinical trials that sampled moderate-to-high-risk patient populations. All costs were reported in US dollars ($).
Analysis of uncertainty:
One-way, two-way, and probabilistic sensitivity analyses were performed. The results were presented in a tornado diagram and a cost-effectiveness acceptability curve. Scattergrams were also generated from 100,000 Monte Carlo simulations.