Modification of empiric antibiotic treatment in patients with pneumonia acquired in the intensive care unit. ICU-Acquired Pneumonia Study Group.

OBJECTIVE: To assess the frequency of and the reasons for changing empiric antibiotics during the treatment of pneumonia acquired in the intensive care unit (ICU).

DESIGN: A prospective multicenter study of 1 year's duration.

SETTING: Medical and surgical ICUs in 30 hospitals all over Spain.

PATIENTS: Of a total of 16,872 patients initially enrolled into the study, 530 patients developed 565 episodes of pneumonia after admission to the ICU.

RESULTS: Empiric antibiotics were administered in 490 (86.7%) of the 565 episodes of pneumonia. The antimicrobials most frequently used were amikacin in 120 cases, tobramycin in 110, ceftazidime in 96, and cefotaxime in 96. Monotherapy was indicated in 135 (27.6%) of the 490 episodes, a combination of two antibiotics in 306 episodes (62.4%), and a combination of three antibiotics in 49 episodes (10%). The empiric antibiotic treatment was modified in 214 (43.7%) cases because of isolation of a microorganism not covered by treatment in 133 (62.1%) cases, lack of clinical response in 77 (36%), and development of resistance in 14 (6.6%). Individual factors associated with modification of empiric treatment identified in the multivariate analysis were microorganism not covered (relative risk (RR)) 22.02; 95% confidence interval (CI) 11.54 to 42.60; p < 0.0001), administration of more than one antimicrobial (RR 1.29; 95% CI 1.02 to 1.65; p = 0.021), and previous use of antibiotics (RR 1.22; 95% CI 1.08 to 1.39; p = 0.0018). Attributable mortality was 16.2% in patients with appropriate initial therapy and 24.7% in patients with inappropriate treatment (p = 0.034).

CONCLUSIONS: A high percentage of patients (43.7%) required modification of empiric antibiotic treatment for pneumonia acquired in the ICU. In 62.1% of cases the main reason for changing antibiotic treatment was inadequate antibiotic coverage of microorganisms. Attributable mortality was significantly higher in patients with inappropriate initial antibiotic therapy. Rapid and accurate diagnostic methods are needed to initiate appropriate antibiotic treatment as soon as pneumonia is suspected.