Cost-Effectiveness of Alternative Anticoagulation Strategies for Postoperative Management of Total Knee Arthroplasty Patients

Abstract

Objective: To evaluate the cost-effectiveness of prolonged (35-day) and standard-duration (14-day) anticoagulation therapy following total knee arthroplasty (TKA).

Methods: Using Markov modeling, we assessed clinical and economic outcomes of 14-day and 35-day anticoagulation therapy following TKA with rivaroxaban, low molecular weight heparin (LMWH), fondaparinux, warfarin, and aspirin. Incidence of complications of TKA and anticoagulation therapy (deep vein thrombosis [DVT], pulmonary embolism [PE], prosthetic joint infection [PJI], and bleeding) were derived from published literature. Daily costs ranged from $1 (aspirin) to $43 (fondaparinux). Primary outcomes included quality-adjusted life years (QALYs), direct medical costs, and incremental cost-effectiveness ratios (ICERs) at 1 year post-TKA. The preferred regimen was the regimen with highest QALYs maintaining an ICER below the willingness-to-pay threshold ($100,000/QALY). We conducted probabilistic sensitivity analyses, varying complication incidence and anticoagulation efficacy, to evaluate the impact of parameter uncertainty on model results.

Results: Aspirin resulted in the highest cumulative incidence of DVT and PE, while prolonged fondaparinux led to the largest reduction in DVT incidence (15% reduction compared to no prophylaxis). Despite differential bleeding rates (ranging from 3% to 6%), all strategies had similar incidence of PJI (1% to 2%). Prolonged rivaroxaban was the least costly strategy ($3,300 at 1 year post-TKA) and the preferred regimen in the base case. In sensitivity analyses, prolonged rivaroxaban and warfarin had similar likelihoods of being cost-effective.

Conclusion: Extending postoperative anticoagulation therapy to 35 days increases QALYs compared to standard 14-day prophylaxis. Prolonged rivaroxaban and prolonged warfarin are most likely to be cost-effective post-TKA; the costs of fondaparinux and LMWH precluded their being preferred strategies.

Figure 1

This figure depicts the model structure used to assess anticoagulation strategies after total knee arthroplasty (TKA). Following TKA, subjects can experience deep vein thrombosis, prosthetic joint infection, or bleeding or can proceed with no complications. Those who experience a complication and do not die enter a post-complication state, which are specific to the complication experienced (post-DVT, post-PE, post-bleed, and post-PJI). Subjects are at continued risk of deep vein thrombosis in the post-complication state. Death can occur at any point.

Figure 2

The total quality-adjusted life days and proportion of the first post-operative year spent in each health state is depicted for each prolonged strategy. Abbreviations: DVT, deep vein thrombosis, PE, pulmonary embolism; PJI, prosthetic joint infection; LMWH, low molecular weight heparin.

Figure 3

The individual components of the annual cost for each prolonged strategy are shown. Anticoagulation includes the cost of the anticoagulant and any monitoring (warfarin) or administration (LMWH and fondaparinux), as appropriate. The cost of anticoagulation in the prolonged aspirin strategy is minimal and not indicated here. Abbreviations: DVT, deep vein thrombosis, PE, pulmonary embolism; PJI, prosthetic joint infection; LMWH, low molecular weight heparin.

Figure 4

This figure shows the proportion of iterations where a given strategy was the cost-effective option at various WTP thresholds. Strategies with probabilities of cost-effectiveness <5% are not shown. Abbreviations: WTP, willingness to pay; LMWH, low molecular weight heparin.