OBJECTIVES: Thrombocytopenia is a serious complication that may occur in patients receiving venous thromboembolism prophylaxis for a variety of reasons, including drug-induced thrombocytopenia. This analysis was conducted to assess if differences exist in the coded rates of thrombocytopenia in patients receiving fondaparinux, dalteparin, enoxaparin, or unfractionated heparin (UFH). METHODS: This was a retrospective analysis of inpatient data from >500 hospitals in the United States. Patients hospitalized for hip or knee replacement or hip fracture surgery between 1/2003 and 3/2005 were eligible for study inclusion. Patients receiving fondaparinux, dalteparin, enoxaparin or UFH <2 days after surgery were included. A control group of orthopedic surgery patients that did not receive any of the anticoagulants of interest was also identified. Patients <18 yrs of age or those receiving >1 anticoagulant of interest during their hospitalization were excluded. The occurrence of thrombocytopenia was determined by the presence an ICD-9 code for thrombocytopenia unspecified (287.5) and/or secondary thrombocytopenia due to circulating anticoagulants (287.4 with E934.2). Logistic regression models were used to assess differences in thrombocytopenia rates; controlling for age, gender, severity of illness (Charlson), length of stay, number of hospitalizations prior to index visit, type of orthopedic surgery, and cancer diagnosis. RESULTS: A total of 250,600 patients were included in the analysis: fondaparinux=11,633; dalteparin=14,713; enoxaparin=92,776; UFH=18,904; control=112,574. The unadjusted rates of thrombocytopenia in each cohort were: fondaparinux=0.8%, dalteparin=1.2%, enoxaparin=1.4%, UFH=1.3%, control=0.9%. After controlling for baseline covariates, patients on fondaparinux were least likely to experience thrombocytopenia. The odds of thrombocytopenia for each anticoagulant when compared to control were: fondaparinux OR=0.98, p=0.83; dalteparin OR=1.2, p=0.02; enoxaparin OR=1.3, p<0.0001; UFH OR=1.2, p=0.03. CONCLUSIONS: Although the relative rates of drug-induced thrombocytopenia cannot be defined with certainty (coding limitations), the risk of thrombocytopenia in fondaparinux-treated patients was similar to control, while patients receiving heparins experienced an increased risk.