RESUMO
OBJECTIVE: The authors had for aim to make an inventory of antibiotic treatment for severe community-acquired Streptococcus pneumoniae pneumonia and compare local practices to the local and national guidelines. PATIENTS AND METHOD: An audit was conducted retrospectively in the Versailles hospital ICU between January 2006 and April 2009. Forty patients were included. RESULTS: Ninety-three percent had major risk factors for pneumonia. Ninety-eight percent were treated, with the usual empirical treatment (69%) or treatment active against Pseudomonas aeruginosa (31%). Eighty-five percent of empirical treatment complied with the French national guidelines issued by the SPILF and 49% with the local ICU protocol, more restrictive for the choice of the agent and dose. Early de-escalation to amoxicillin was applied to 41% of patients after obtaining results for pneumococcal and Legionella antigen and results of respiratory sample direct examination. For all patients, empirical treatment was reassessed according to culture results: 81% were prescribed amoxicillin. Evaluation showed that 92% of treatment complied with SPILF guidelines and 65% with the local ICU protocol that required adaptation of amoxicillin doses according to MICs; adaptation to severity and BMI was necessary for ten patients. Mortality remained high, at 37%, despite using antibiotics still effective against S. pneumoniae. CONCLUSIONS: This survey revealed a satisfactory adhesion to recommendations and prompt responsiveness of the team for adjustment of antibiotic therapy. The audit allowed updating the local ICU protocol.