Purpose: Central venous lines are the most important risk factor for pediatric deep venous thromboses (DVTs). Management of pediatric catheter-associated DVTs is quite variable. In the absence of clinical trials, we evaluated whether treating catheter-related DVTs with enoxaparin is more effective than no treatment.

Methods: Using DECISION MAKERŪ we constructed a decision analytic model quantifying risks and benefits of treatment of pediatric catheter-related DVTs with enoxaparin for 6 months. Data informing the model were obtained from English language literature identified through MEDLINE searches. We considered short-term morbidities including pulmonary embolism (PE), major bleeding, and death from these events. Long-term morbidities included post-thrombotic syndrome (PTS) and the sequelae associated with major bleeding. We also considered the disutility associated with prolonged hospitalization and the need for subcutaneous anticoagulation. Effectiveness was measured in quality-adjusted life years (QALYs). Uncertainty in parameter values was explored through sensitivity analyses. The base case focused on a 3 mo full term infant with a catheter-related DVT.

Results: Treatment with enoxaparin yielded a slightly higher expected utility than no treatment (74.84 vs 74.63 QALYs). Sensitivity analysis on the probability of bleeding demonstrated that enoxaparin is favored at probabilities below 4.3%; this threshold is within a plausible clinical range according to the literature (0-4.9%). PE-associated mortality was 11% in the base case; the enoxaparin strategy dominated above a 30-day cumulative mortality of 7.4% (literature range: 0-17%). The probability of PE without treatment is unknown (base case value 25%), although the enoxaparin strategy dominated above a probability of 16.9%. Within clinically plausible ranges of probabilities the model was insensitive to parameters that were large sources of uncertainty: probability of PE without treatment, probability of post-thrombotic syndrome, probability of dying of major bleeding, long-term quality of life after major bleeding, efficacy of enoxaparin, and quality of life after major bleeding.

Conclusions: Using decision analytic modeling, we found that treatment with enoxaparin yields only a small gain in quality-adjusted life expectancy compared with no treatment of pediatric catheter-associated DVTs. Therefore, risks of anticoagulation in children appear to be balanced by the risks of pulmonary embolism.