BACKGROUND: Patients receiving heparin for thromboprophylaxis or treatment may have new or recurrent venous thromboembolism (VTE) if immune-mediated heparin-induced thrombocytopenia (HIT) occurs or for other reasons, eg, if anticoagulation fails. We estimated from the literature how frequently a patient presenting with VTE during or following heparin therapy has HIT-associated VTE. METHODS: A comprehensive, systematic literature search was conducted to identify studies using unfractionated or low-molecular-weight heparin (LMWH) for thromboprophylaxis or treatment in which new or recurrent VTE and serologically confirmed HIT were reported. From extracted study data, the proportion of patients with HIT-associated VTE relative to any VTE was calculated by heparin type and mode of administration. RESULTS: We identified 10 studies, some with multiple arms, that used unfractionated heparin (IV administration, 5 studies; subcutaneous administration, 3 studies) or subcutaneous LMWH (5 studies) and met analysis criteria. Across these studies, 386 of 6,219 heparin-treated patients had VTE, including 32 patients who also had HIT. The frequency of HIT-associated VTE among heparin-treated patients with VTE was comparable between IV and subcutaneous unfractionated heparin therapy (13.2% [17 of 129 patients] vs 12.4% [14 of 113 patients]; odds ratio, 1.07; 95% confidence interval, 0.50 to 2.3; p > 0.99) yet significantly different between unfractionated heparin and LMWH therapy (12.8% [31 of 242 patients] vs 0.7% [1 of 144 patients]; odds ratio, 21.0; 95% confidence interval, 2.8 to 156; p < 0.001). CONCLUSIONS: VTE is associated with HIT infrequently (< 1%) in LMWH-treated patients, yet often (approximately one in eight cases) in unfractionated heparin-treated patients. Physicians should suspect the possibility of HIT if VTE develops during or soon after unfractionated heparin use; if thrombocytopenia is present, alternative anticoagulation should be used until HIT is excluded.