PURPOSE: In cases of suspected infection, empiric, broad spectrum antibiotics are often initiated to prevent worsening infections, severe sepsis, multiorgan failure, and death. It is widely accepted that antibiotic de-escalation should be promptly performed once bacterial sensitivities have been identified to reduce rates of antimicrobial resistance. The goal of our study was to determine outcomes associated with appropriate antibiotic management strategies, including de-escalation in critically ill patients.
METHODS: A retrospective review of 39 cases of bacteremia occurring during July –December 2007 in critically ill patients admitted to our facility's medical intensive care unit (MICU) was performed. We focused on benchmarks of antimicrobial management strategies, specifically time to initial antibiotics, time to appropriate antibiotics, and time to de-escalation of antibiotics. In addition, we analyzed several outcomes, including hospital and ICU length of stay (LOS), defervescence, reversal of leukocytosis, and mortality. A database was created, and statistical analysis was performed using SPSS, focusing on comparison of means, as well as survival analysis when appropriate.
RESULTS: When analyzing mortality outcomes, there was a statistically significant difference in age (48.59 vs 63.95, p-value 0.009) and time to initial antibiotics (1.76 vs 13.13, p-value 0.013)when comparing non-survivors vs survivors. There was also a trend to significance with respect to race, with higher mortality seen in African Americans (p=0.065). There were no statistically significant associations between antibiotic de-escalation and 1) hospital LOS, 2) MICU LOS, 3) time to defervescence, and 4) mortality.
CONCLUSION: The antibiotic management strategy of bacteremic patients was no different in patients that lived or died. Despite similar severity of illness by APACHE II scores, younger age and shorter time to antibiotics were associated with increased mortality. These relationships were unexpected, and should be clarified by larger studies. Underlying causes further explaining survival differences between races may require further investigation.
CLINICAL IMPLICATIONS: Age may influence survival in critically ill patients with bacteremia. Questions about race and outcomes may warrant further studies.
DISCLOSURE: Elizabeth Lechner, None.