To compare the cost effectiveness of low-molecular-weight heparin with that of oral anticoagulants in preventing recurrences after an episode of venous thromboembolism.

A decision tree was used to assess the cost and the expected quality-adjusted years of life (QALY) after treatment with either low-molecular-weight heparin or warfarin, based on pooled data from six published trials. Preferences were elicited with a modified time trade-off method in a sample of patients attending an anticoagulation clinic.

Compared with warfarin, low-molecular-weight heparin significantly decreased the rate of minor bleeding (odds ratio [OR] = 0.24; 95% confidence interval [CI]: 0.14 to 0.43) but not recurrent deep vein thromboses (OR = 0.77; 95% CI: 0.43 to 1.35). Patients’ preference for warfarin (0.988, on a 0 to 1 scale) was lower than that for low-molecular-weight heparin (0.992), but the difference was not statistically significant. A Monte Carlo analysis estimated that low-molecular-weight heparin saved an average of 13 quality-adjusted days compared with warfarin, at a cost of $6,583 per QALY (95% CI: $5,525 to $7,625) based on costs in Italy and $28,231 per QALY (95% CI: $20,872 to $36,773) based on costs in the United States. When we included rebound recurrences after interruption of therapy, which were more common with low-molecular-weight heparin, treatment with low-molecular-weight heparin cost $53,166 per QALY in Italy and $177,166 per QALY in the United States.

Low-molecular-weight heparin might be a cost-effective drug for secondary prophylaxis of venous thromboembolism, especially in patients at high risk of recurrence and where the drug’s cost is lower. The apparent increase in recurrence after interruption of therapy needs to be investigated more thoroughly before low-molecular-weight heparin can be recommended routinely.