Analytical approach:
A Markov decision analytic model was used to simulate the costs and outcomes of the two interventions. The time horizon was the lifetime of a hypothetical cohort of 46-year-olds. The perspective was societal.
Effectiveness data:
Clinical and effectiveness data were derived from published vascular surgery and trauma literature. The authors reported that for data unavailable from the literature, published evidence using similar populations was used to make assumptions. The main measure of effectiveness used in the study was the probability of death or venous thromboembolism. These estimates were derived from previously published studies.
Monetary benefit and utility valuations:
The authors reported that utilities were derived from the published literature and New England Medical Centre Utility Search Database.
Measure of benefit:
Quality-adjusted life-years (QALYs) gained. Discounted using an annual rate of 3%.
Cost data:
Direct costs included costs for hospitalisations (including complications), prophylactic and therapeutic anticoagulation, computed tomography of the thorax, filter placement and complications, filter removal and treatment of venous insufficiency. Costs were derived from the Medicare and Medicaid Services fee schedules, Healthcare Cost and Utilisation Project (HCUP) database and the Red Book for wholesale drug prices. Indirect costs included those due to lost wages due to hospitalisation and treatment. Time off work was valued by the authors at the hourly private non-farm worker wage. Cost of death was valued at $5,000. All costs were reported in 2007 US dollars ($) and were discounted using an annual rate of 3%.
Analysis of uncertainty:
The authors reported that they performed a series of one-way sensitivity analyses to assess how robust the base case results were and determine the most influential variables.