Universal Thrombophilia Screening Not Cost-Effective for Determining Duration of Anticoagulation in Pediatric Patients.

Background: The use of thrombophilia screening to determine anticoagulation duration in children with deep venous thrombosis (DVT) remains controversial. The American College of Chest Physicians guidelines for antithrombotic therapy in children recommend 3 months of anticoagulation for DVT, with longer treatment for idiopathic DVTs or the presence of ongoing risk factors. Cost-effectiveness analyses in adults have shown that thrombophilia screening followed by prolonged anticoagulation in screen-positive patients may be cost-effective. We designed a decision analytic model addressing health effects and costs of universal thrombophilia screening in pediatric patients with a first episode of DVT.

Methods: Our 2-year Markov model evaluated the cost-utility of three management strategies:

1. no screening, anticoagulate for 3 months,

2. screening, anticoagulate unaffected patients for 3 months, affected patients for 6 months, and

3. no screening, anticoagulate for 6 months.

We performed a literature search to estimate prevalence of thrombophilia in children with DVT, risk of DVT recurrence and major bleeding, efficacy of anticoagulation, and DVT-related and all-cause mortality. We obtained quality of life measures (utilities) and cost data from published sources. Costs included hospitalization, medications, thrombophilia screen (Factor V Leiden, prothrombin 20210A, homocysteine, protein C and S, antithrombin, homocysteine, antiphospholipid antibodies) and other laboratory monitoring, ultrasounds, physician visits, and lost parental wages. With the exception of antiphospholipid antibodies, the influence of thrombophilia on recurrent DVT risk is controversial. We used a hazard ratio of 2.4 for DVT recurrence in affected patients, but varied this widely in sensitivity analysis. Using a societal perspective, we compared strategies based on costs and quality-adjusted life-years (QALYs).

Results: Total strategy costs were $7500 for no screen, treat for 3 months; $8500 for screen, then treat for 3 or 6 months; and $11,700 for no screen, treat for 6 months. Three months treatment without screening was not only the least expensive strategy but also the most effective (1.74 QALYs) by 0.01–0.03 QALYs when compared to other strategies. No screening with 3 months anticoagulation becomes less effective than other strategies only when the probability of DVT recurrence is >60% per month in patients with thrombophilia (baseline estimate from the literature=0.012%) or >45% per month in patients without thrombophilia (baseline=0.005%). Incremental cost-utility ratios were sensitive to variation in hospitalization and medication costs, but 3 months, no screening always remained the preferred choice.

Discussion: Our results demonstrate that universal thrombophilia screening is not a cost-effective strategy when used to determine anticoagulation duration in pediatric patients unless predicted DVT recurrence rates are extraordinarily high. Thrombophilia screening has other potential benefits (prophylaxis in future high-risk settings, screening family members). However, a full panel of tests does not need to be a “knee-jerk” reaction at the time of every DVT diagnosis. Rather, for many patients, screening can take place at a later date on a case-by-case basis, with patient and family involvement in the decision.