Human immunodeficiency virus infection and risk of venous thromboembolism.

BACKGROUND: The incidence of venous thromboembolism (VTE) in HIV-infected patients is uncertain, and the impact of protease inhibitors on the incidence of VTE is also uncertain, yet important to know to create a database for providing opinion regarding prophylaxis for the prevention of VTE.
METHODS: Data from the National Hospital Discharge Survey (NHDS) were analyzed from 1990 through 2005. International Classification of Diseases, Ninth Revision, and Clinical Modification (ICD-9-CM) codes were used to identify illnesses.
RESULTS: Among 2,429,000 patients older than 18 years hospitalized with HIV infection, the incidence of pulmonary embolism (PE) was 0.4%, deep venous thrombosis (DVT) 1.4%, and VTE 1.7%. The relative risks compared with all hospitalized non-HIV patients of PE, DVT, and VTE were 0.91, 1.26, and 1.21. The incidence of VTE from 1990 to 1996 was 17,000 of 1,198,000 (1.4%) and after 1996 it was 25,000 of 1,230,000 (2.0) (P < 0.0001) (relative risk = 1.43). Among hospitalized patients who did not have HIV, the relative risk comparing incidences after 1996 with incidences before was 1.22.
CONCLUSION: The incidence of VTE in patients with HIV infection was higher than in non-HIV patients. The incidence of VTE in patients with HIV in the postprotease inhibitor era (after 1996) was higher than in HIV patients before 1996, but the incidence was also higher in non-HIV patients after 1996. The higher incidence since 1996 is small, probably not clinically significant, and not necessarily because of protease inhibitors.