OBJECTIVES: Heparin-induced thrombocytopenia (HIT) is a prothrombotic condition that requires immediate treatment, usually with direct thrombin inhibitors (DTIs). However, the treatment may   cause major bleeding. Data comparing the effectiveness and safety of DTIs in the treatment of HIT is limited. Therefore, we compared the effectiveness and safety of two commonly used DTIs, argatroban and bivalirudin, in patients with suspected HIT in a real-world setting. METHODS: The Clinical Database/Resource Manager data from the UHC were used to identify hospitalized patients with suspected HIT discharged between 2009 to 2012. Using propensity scores, up to three argatroban-treated patients were matched to each bivalirudin-treated patient. The rates of thrombosis, major bleeding, amputation and mortality were compared using hazard ratios (HRs) from Cox proportional hazard regression models. RESULTS: Of the 2,408 matched patients, 709 (29.4%) received bivalirudin and 1,699 (70.6%) received argatroban. Compared to bivalirudin, argatroban exhibited a less harmful effect for major bleeding based on blood transfusion (6.1% vs. 11.4%; P<0.05). The result did not change when major bleeding was defined as blood transfusion with diagnostic codes.  The risk of thrombosis (7.7% vs. 9.5%; P=0.14), amputation (2.1% vs. 2.0%; P=0.82) and mortality (24.2% vs. 25.5%; P=0.49) were similar in argatroban-and bivalirudin-treated patients. CONCLUSIONS: Among patients with suspected HIT, the use of argatroban compared to bivalirudin was associated with a decreased risk of major bleeding but was similar for thrombosis, amputation and mortality risk.