| To Register SMDM Homepage | |
|
|
|
||||
Objective: To use a Bayesian approach to evaluate the cost-effectiveness of enoxaparin (ENOX) versus unfractionated heparin (UH) for the prophylaxis of deep-vein thrombosis (DVT) following major trauma. Methods: A decision analytic model was used to measure the incremental cost and incremental effectiveness of ENOX vs. UH as the comparator for DVT prophylaxis from the hospital perspective. Outcomes data were extracted from the only published clinical trail (NEJM 1996; 335 (10):701-7). The probability of death from a major bleed or pulmonary embolism (PE), and additional model parameters were derived from published data and the Ontario Trauma registry, with hospital costs derived from the Ontario case-costing project. The rates of venographically detected DVT were adjusted for the sensitivity and specificity of clinically and ultrasonographically diagnosed DVT. The primary outcome measures were the incremental cost per DVT averted and cost per life-year gained. Probabilistic sensitivity analysis was performed using 2nd order Monte Carlo simulation. All costs are in 2003 CDN$. Results: The total cost of treatment with enoxaparin was $13,395 versus $13,229 with UH, resulting in an incremental cost of $97. Enoxaparin resulted in an incremental effect of 0.085 DVTs averted and -0.29 life years gained. These results yielded an incremental cost effectiveness ratio of $1,139 per DVT averted however, when life years gained was used as the metric of effectiveness, enoxaparin was dominated by UH. Monte Carlo simulation revealed that in the DVTs averted model, 98% of the model iterations fell in the NE and SE quadrants, favoring ENOX. Conversely, in the life years gained model, 96% of the model iterations fell in the NW and SW quadrants, favoring UH. At λ=$70,000 per life year, there was only a 5% probability that enoxaparin was cost effective. Conclusions: Only one previous study has evaluated the cost-effectiveness (cost/DVT averted) of DVT prophylaxis in this clinical scenario which showed that ENOX was the dominant strategy. However, neither the costs or outcomes related to PE, or mortality related to PE or major bleed were incorporated into the model. This study demonstrates the importance of considering mortality when modeling DVT prophylaxis-related outcomes in the trauma population, and the benefits of a Bayesian modeling approach to the analysis.
See more of Poster Session - CEA: Methods and Applications; Health Services Research
See more of The 26th Annual Meeting of the Society for Medical Decision Making (October 17-20, 2004)