Rifampin use in acute community-acquired meningitis in intensive care units: the French retrospective cohort ACAM-ICU study

Cédric Bretonnière, Mathieu Jozwiak, Christophe Girault, Pascal Beuret, Jean-Louis Trouillet, Nadia Anguel, Jocelyne Caillon, Gilles Potel, Daniel Villers, David Boutoille, Christophe Guitton

Bacterial meningitis among critically ill adult patients remains associated with both high mortality and frequent, persistent disability. Vancomycin was added to treatment with a third-generation cephalosporin as recommended by French national guidelines. Because animal model studies had suggested interest in the use of rifampin for treatment of bacterial meningitis, and after the introduction of early corticosteroid therapy (in 2002), there was a trend toward increasing rifampin use for intensive care unit (ICU) patients. The aim of this article is to report on this practice. Five ICUs participated in the study. Baseline characteristics and treatment data were retrospectively collected from charts of patients admitted with a diagnosis of acute bacterial meningitis during a 5-year period (2004-2008). The ICU mortality was the main outcome measure; Glasgow Outcome Scale and 3-month mortality were also assessed. One hundred fifty-seven patients were included. Streptococcus pneumoniae and Neisseria meningitidis were the most prevalent causative microorganisms. The ICU mortality rate was 15 %. High doses of a cephalosporin were the most prevalent initial antimicrobial treatment. The delay between admission and administration of the first antibiotic dose was correlated with ICU mortality. Rifampin was used with a cephalosporin for 32 patients (ranging from 8 % of the cohort for 2004 to 30 % in 2008). Administration of rifampin within the first 24 h of hospitalization could be associated with a lower ICU survival. Statistical association between such an early rifampin treatment and ICU mortality reached significance only for patients with pneumococcal meningitis (p=0.031) in univariate analysis, but not in the logistic model. We report on the role of rifampin use for patients with community-acquired meningitis, and the results of this study suggest that this practice may be associated with lower mortality in the ICU. Nevertheless, the only independent predictors of ICU mortality were organ failure and pneumococcal infection. Further studies are required to confirm these results and to explain how rifampin use would reduce mortality.