[Antimicrobial blood concentrations in burns. A five years' retrospective survey]

The monitoring of antimicrobial agents is a routine in our unit. We reviewed the results either of peak-and-through concentrations (peak and through is one sample) or concentrations at steady state (Css) of all antimicrobials given over five years (2001 to 2005) and studied the antimicrobials with at least 20 samples. We found 706 samples in 122 patients, the antimicrobials being amikacin, amoxicillin, ceftazidime, ciprofloxacin, cloxacillin, gentamicin, imipenem, ofloxacin, tobramycin and vancomycin. When comparing samples in witch the concentrations were above a value thought to be predictive of efficacy to those were not, we could notice: that no parameter about patients or burn surface was predictable for achieving targeted blood concentration; that usual regimen could not achieve targeted concentrations, excepted with ceftazidime, provided it was used in continuous infusion; that, with the other beta-lactams, continuous infusion was more likely to achieve targeted blood concentrations; that, with fluoroquinolones, both higher and more frequent injections were needed; that, with aminoglycosides used once a day, the dosage had to be higher than usually recommended. We conclude that antimicrobial regimen should be altered in burns and that a monitoring of blood concentrations should be performed in these patients.