Major but differential decline in the incidence of Staphylococcus aureus bacteraemia in HIV-infected individuals from 1995 to 2007: a nationwide cohort study*.

OBJECTIVES: Incidence rates (IRs) of Staphylococcus aureus bacteraemia (SAB) are known to be higher in HIV-infected individuals than in the general population, but have not been assessed in the era of highly active antiretroviral therapy. METHODS: From 1 January 1995 to 31 December 2007, all Danish HIV-infected individuals (n=4871) and population controls (n=92 116) matched on age and sex were enrolled in a cohort and all cases of SAB were registered. IRs and risk factors were estimated using time-updated Poisson regression analysis. RESULTS: We identified 329 cases of SAB in 284 individuals, of whom 132 individuals were infected with HIV and 152 were not [crude IR ratio (IRR) 24.2; 95% confidence interval (CI) 19.5-30.0, for HIV-infected vs. non-HIV-infected individuals]. Over time, IR declined for HIV-infected individuals (IRR 0.40). Injecting drug users (IDUs) had the highest incidence and the smallest decline in IR, while men who have sex with men (MSM) had the largest decline over time. Among HIV-infected individuals, a latest CD4 count <100 cells/µL was the strongest independent predictor of SAB (IRR 10.2). Additionally, HIV transmission group was associated with risk of SAB. MSM were more likely to have hospital-acquired SAB, a low CD4 cell count and AIDS at the time of HIV acquisition compared with IDUs. CONCLUSIONS: We found that the incidence of SAB among HIV-infected individuals declined during the study period, but remained higher than that among HIV-uninfected individuals. There was an unevenly distributed burden of SAB among HIV transmission groups (IDU>MSM). Low CD4 cell count and IDU were strong predictors of SAB among HIV-infected individuals.

Mortality in enterococcal bloodstream infections increases with inappropriate antimicrobial therapy.

Enterococcus species are common in nosocomial bloodstream infections and their incidence is rising. Although well recognized in several serious bacterial infections, the influence of appropriate antimicrobial therapy in enterococcal bacteraemia has not been fully settled. The aim of the study was to determine whether administration of inappropriate antibiotics in enterococcal bacteraemia is an independent risk factor for mortality, among other known and suspected risk factors. We conducted a cohort study of E. faecalis/faecium bacteraemia during a 3-year period at a single tertiary care hospital in Denmark. Patients with growth of non-enterococcus co-pathogens apart from the enterococcal bacteraemia were also included, as were patients with repeated enterococcal bacteraemia. Time to appropriate antimicrobial therapy was counted from the first episode. Appropriate antibiotic therapy was defined as any therapy with documented clinical effect, in vitro activity and a minimum treatment length of 6 days. Multivariate regression models were built to determine the independent risk factors for mortality. We included 196 patients with enterococcal bacteraemia. Appropriate antibiotics for at least 6 days were administered in 146 of these (74%). Thirty-day mortality was 26%. Multivariate logistic regression identified independent predictors of 30-day all-cause mortality: appropriate antimicrobial therapy for ≥ 6 days (odds ratio for mortality 0.33, 0.14-0.79), ICU admission (4.2, 1.7-10), thrombocytopenia (3.9, 1.6-9.3), chronic liver failure (3.3, 1.1-10) and age ≥ 60 years (2.2, 0.99-5.0). Antibiotics not appropriately covering enterococci are frequently administered empirically in suspected bloodstream infections. Inappropriate antibiotic therapy was an independent risk factor for mortality in enterococcal bacteraemia.