Epidemiology of thrombosis in HIV-infected individuals

Objective: To describe the incidence of clinically recognized thrombosis and associated factors among individuals infected with HIV. Design: A longitudinal medical record review. Setting: Over 100 medical clinics in nine US cities. Patients: A total of 42 935 individuals aged 13 years or older with HIV infection, observed for an average of 2.4 years. Main outcome measures: The incidence of thrombosis among HIV-infected individuals; adjusted odds ratios for factors associated with thrombosis. Results: The incidence of thrombosis among HIV-infected individuals was 2.6/1000 person-years (PY). Factors significantly associated with thrombosis included: age of 45 or more years (adjusted odds ratio [AOR], 1.9; 95% confidence interval [CI], 1.4-2.7); a diagnosis of cytomegalovirus disease or retinitis (AOR, 1.9; Cl, 1.2-2.9), or other AIDS-defining opportunistic illness (AOR, 1.5; CI, 1.1 -2.2); hospitalization (AOR, 3.3; CI, 2.5-4.4); and the prescription of megestrol acetate (AOR, 2.0; CI 1.3-2.9) or indinavir (AOR, 2.4; Cl 1.4-4.3). The prescription of other protease inhibitors, sex, race, and mode of HIV exposure were not associated with thrombosis. Conclusion: Among HIV-infected individuals, clinically detected thrombosis is more common in those who have opportunistic illnesses, for whom megestrol acetate or indinavir have been prescribed, who have been hospitalized, and who are aged 45 years or older. Physicians should be aware of the risks of thrombosis in order to promote the early identification and appropriate treatment or prophylaxis. Further study is needed to characterize the association between indinavir and thrombosis. (C) 2000 Lippincott Williams & Wilkins.