A decision-tree analysis was used to estimate the average cost per patient using the direct thrombin inhibitor argatroban for early treatment (<48 hours after thrombocytopenia onset) compared with delayed treatment (> or =48 hours after thrombocytopenia onset) of immune-mediated heparin-induced thrombocytopenia (HIT) with or without thrombosis. Clinical probability data used to populate the model were obtained from argatroban clinical trials and from published clinical literature. Resource utilization data and cost data were also obtained from available literature, the 2003 Physician's Fee Reference, the Healthcare Cost and Utilization Project 2000, the 2003 Drug Topics RedBook, and a modified Delphi panel. The total per-patient cost included hospital days, diagnostic tests, heparin, argatroban, major hemorrhagic events, and patient outcomes (ie, amputation, new thrombosis, stroke, or death), multiplied by the probability of each event. The incremental cost-effectiveness ratio was calculated by dividing the incremental cost between patients with and without argatroban treatment by the incremental effectiveness, or the cost per new thrombosis event avoided. The mean cost per HIT patient without thrombosis who did not receive argatroban was $38,046. The mean cost decreased by 6.85% for patients who were treated earlier with argatroban therapy (average cost, $35,441), representing a $2605 saving per patient compared with those not treated with argatroban. For those receiving delayed argatroban therapy, the mean cost increased by $9024 per patient compared with those receiving early treatment with argatroban. The mean cost for HIT patients with thrombosis who did not receive argatroban was $48,101, which was 9.0% higher than for those receiving early argatroban therapy, representing a $3957 savings per patient. For HIT with thrombosis, mean costs increased by 18.2% in patients whose argatroban was delayed, representing a cost increase of $8020 per patient compared with early treatment (mean cost $44,144 for early treatment and $52,164 for delayed treatment). The results of this analysis support the recommendation to initiate early argatroban treatment upon suspicion of HIT to reduce the thrombotic consequences of HIT and associated healthcare costs. Argatroban therapy should not be delayed pending the results of HIT diagnostic tests.