Background:
In human immunodeficiency virus (HIV) disease, neutropenia occurs most commonly in patients who are also severely immunosuppressed. It is not currently known whether neutropenia is an independent risk factor for the development of bacterial infection, which is a potentially serious complication of advanced HIV disease.
Methods:
We compared the incidence of bacterial infection between 118 neutropenic patients (absolute neutrophil count [ANC], <1×109/L) and 118 nonneutropenic patients matched for CD4+ lymphocyte count, use of injecting drugs, and follow-up time from a demographically heterogeneous urban cohort of HIV-infected patients followed up longitudinally at the Johns Hopkins Hospital. The incidence of serious infection was analyzed separately for patients with an ANC of less than 1, less than 0.75, or less than 0.5×109/L.
Results:
There were no statistically significant associations found between neutropenia and several individual bacterial infections, including bacteremia, pneumonia, endocarditis, bacterial enterocolitis, and infection of normally sterile sites for any level of neutropenia. However, for all these infections combined, the adjusted relative risk for the occurrence of bacterial infection was 2.33 (95% confidence interval, 1.00 to 5.40; P=.05) for patients with an ANC of less than 1×109/L and 7.92 (95% confidence interval, 1.18 to 53.2; P=.03) for those with an ANC of less than 0.5×109/L. The incidence of serious bacterial infection ranged from two to three infections per 100 person-months of neutropenia for patients with an ANC of less than 1×109/L and three to five infections per 100 person-months of neutropenia for patients with an ANC of less than 0.5×109/L for all bacterial infections combined.
Conclusions:
Our matched cohort analysis indicates that neutropenia is an independent risk factor for bacterial infection in patients with advanced HIV disease. Given the incidence of infection, the cost-effectiveness of interventions to prevent neutropenia in advanced HIV disease should be assessed.(Arch Intern Med. 1995;155:1965-1970)