A prospective study of bloodstream infection in HIV-positive patients during a 15-year period

Abstract number: 1733_145

García Rodríguez J., Álvarez H., Buño B., Mariño A., Rodríguez M., Sesma P.

Objective: To know the evolution of incidence, etiology and prognostic factors for bacteraemia infection in HIV positive patients.

Methods: Prospective study of cases of bacteraemia in our Hospital between 1991 and 2005. Blood cultures were processed using automatic technique and antibiotic susceptibility was determined by microdilution according to the current recommendations from the NCCLS. In patients with HIV positive, demographics as well as a complete set of tests (including clinical features, risk factors to HIV infection, laboratory studies, CD4, treatment and illness evolution) were collected in each case. Statistical analyses were carried out with SPSS.

Results: We studied 1848 episodes of bacteraemia in adult patients, 103 (5.6%) of them in 84 HIV positive patients (incidence by 1000 hospitalised patients-year: 0.62 in pre-HAART and 0.61 in HAART-era). The HIV transmisión were: intravenous drug use 77.7%, heterosexual contact 9.7%, male homosexual contact 2.9%, blood transfusion 1% and unknown 8.7%. Male 87 (84.5%). Mean age 34.2±7 years (range 20–59), 30.4±4.7 in the pre-HAART vs 35.8±7.3 in the HAART-era, p < 0.05. The bacteraemia was community acquired in 82 (79.6%) episodes. 98 (95.1%) episodes were due to monomicrobial infection. The more frequently isolated germs were: Staphylococcus aureus 27 and Streptococcus pneumoniae 14; and the origins: endocarditis 18.4% and pneumonia 30.1%. The CD4 cell count was <50 in 31.4% of the patients, 50–200 in 28.6%, 201–500 in 31.4% and >500 in 8.6%. Among 103 episodes, 17 (16.5%) patients died due to the sepsis and 8 (7.8%) cases were not linked to sepsis. Patients in the HAART-era had more comorbidity (41/72 vs 4/31), more frecuently neutropenia (13/72 vs 1/29), pneumococcal vaccine administration (38/63 vs 4/31)and prophylaxis with TMT-SMX (18/66 vs 2/31) than patients in the pre-HAART era. No statistically significant differences were observed between the pre-HAART and the HAART-era in the isolated germs, origins of bacteraemia, mean CD4 cells count, global mortality (7/31 vs 18/72) nor related to bacteraemia mortality (4/31 vs 13/72).

Conclusions: The incidence, etiology, origin and mortality of bacteraemia in HIV patients have remained stable. Patients in the HAART-era had more frecuently associated comorbility, pneumococcal vaccine administration an prophylaxis with TMT-SMX than those in the pre-HAART era.