Abstract
Our objective was to evaluate the impact of using an imipenem de-escalation protocol for empiric febrile neutropenia on the development of carbapenem resistance. A pre-post intervention design was used. The intervention was adopting the imipenem de-escalation approach, which began on January 1, 2012. A retrospective chart review of cases of febrile neutropenia bacteremia was performed one year before and one year after the intervention. We compared the development of carbapenem resistance between the two study periods. Seventy-five episodes of febrile neutropenia bacteremia were included in the study. They had similar demographics, clinical features and outcomes. There were 78 and 12 pathogens in the primary and follow-up blood cultures, respectively. Approximately 61% and 66% of the primary and follow-up blood cultures, respectively, were gram-negative bacteria with similar carbapenem resistance profiles in the two study periods. In our study population, 57% of the gram-negative bacteria were ESBL pathogens. The resistance of the gram-negative bacteria to piperacillin/tazobactam (72% versus 53%, p=0.161), imipenem (16% versus 11%, p=0.684), and meropenem (8% versus 16%, p=0.638) did not significantly change after our policy change. In conclusion, the use of the carbapenem de-escalation approach for febrile neutropenia in our institution was not associated with an increase in carbepenem resistance. Future prospective multi-center studies are recommended to further confirm the current findings.