Study objective: To determine the incidence, clinical presentation, and risk factors of deep venous thrombosis (DVT) in a high-risk group of ICU patients receiving DVT prophylaxis.

Design: A prospective cohort study.

Setting: Two ICUs of a university-affiliated teaching hospital.

Patients: Patients admitted to the ICUs within 48 h of hospitalization and who had an ICU stay of > or = 4 days underwent venous duplex scans.

Interventions: None.

Results: One hundred two patients were studied. Ninety-four (92%) patients received DVT prophylaxis. Twelve patients (12%) were documented to have DVT by venous duplex scans. There was proximal clot extension in eight of these patients, four of whom had high-probability ventilation/perfusion scans. Of the 56 patients without signs or symptoms of DVT, only two (3.6%) had abnormal scans. Leg swelling was present in 11 patients, six of whom had DVT (p = 0.004). One of 11 patients with unexplained fever had an abnormal scan. Five of the 26 patients (19%) receiving pneumatic compression developed DVT compared with five of 68 patients (7.4%) receiving subcutaneous heparin (not significant). No specific factor was identified that increased the risk of DVT.

Conclusion: In this study, the incidence of DVT in a group of high-risk ICU patients receiving DVT prophylaxis was 12%. Since scans in patients without signs or symptoms suggestive of DVT were abnormal in only 3.6% of patients, venous scans should be performed only in patients with features suggestive of DVT or pulmonary embolism.