Abstract
Background
: Antimicrobial resistance is a growing problem in the intensive care setting. This study was designed to evaluate the trends in bacterial prevalence and changes in antibiotic resistance at a large university hospital over the past decade. Antimicrobial resistance data were compared among the surgical intensive care unit (SICU), medical intensive care unit (MICU), and burn unit (BNU).
Materials and Methods
: A large database was created using hospital-wide data from 1989 to 2000. A retrospective analysis of the evolution of organism prevalence, antibiotic resistance, and response to study protocols was evaluated. The formulary was relatively unrestricted. All positive cultures were examined, focusing on wound, blood, and sputum cultures. Six primary antibiotics were targeted specifically to follow resistance patterns.
Results
: There were 847 identified positive wound cultures, 2,862 positive sputum cultures, and 2,252 positive blood cultures. The incidence of gram-positive and gram-negative organisms changed little in the SICU and BNU; however, there was a large increase in gram-positive organisms and yeast in the MICU over the past 5 years. Anaerobic bacteria and yeast were nearly nonexistent pathogens in the SICU and BNU. The resistance pattern of most organisms changed little following the introduction of a new antibiotic. However, the effectiveness of study antibiotics after formal clinical study periods decreased dramatically, albeit transiently.
Conclusion
: Hospital-wide antibiotic resistance data may be misleading and may not reflect individual critical care units throughout the hospital. Bacterial flora, including resistant organisms, changed little over 10 years, despite an unrestricted formulary. The emergence of resistant and opportunistic organisms is related to antibiotic usage and can vary significantly over time. This suggests that a policy of administering limited duration, narrow spectrum antibiotics may reduce drug resistance.
Get full access to this article
View all access options for this article.