Analysis of data
Outcomes for each intervention (ie, warfarin and no antithrombotic therapy) were analyzed in the decision tree model based on their individual probabilities and costs. Pathway probabilities were multiplied by the sum of the pathway cost to derive an expected cost (Figure 1) for each pathway. Sensitivity analysis was performed on three baseline variables in the warfarin branch (probabilities of minor bleed, major bleed and emboli) to assess the robustness of the data. Break-even analysis was also performed on these variables and the cost of warfarin therapy to determine at which point the two interventions consume identical expected costs. buy ortho tri-cyclen
Expected cost for each intervention was obtained by calculating the sum of expected costs for each outcome. A comparison of expected costs was used to determine the difference between regimens and expected net savings. Cost/ stroke prevented and cost/ death prevented were calculated for warfarin and no antithrombotic therapy. Incremental analysis was also done for cost/ treatment success of stroke and death prevented. Since all data were taken from a single year, discounting was not applied.
Software: Construction and analysis (including sensitivity and break-even analysis) of the decision tree model was done using Decision Analysis by Treeage Software Inc (Massachusetts) (version 2.6).