ABSTRACT
Objective To investigate de-escalation of empiric broad-spectrum antibiotics treatment for patients in intensive care unit (ICU).Methods Data of the patients discharged from ICU in the Second Affiliated Hospital of Zhejiang University from July 1 to December 31 of 2012 and from July 1 to December 31 of 2013 were retrospectively reviewed.Patients with initial use of empirical broad-spectrum antibiotics within 3 d after ICU admission were included in the study.Clinical data including status of infection,the initial empiric antimicrobial therapy,pathogens culture and adjustment of antibiotics in 5 days were analyzed.Results A total of 841 patients were discharged from ICU during the study periods and antibiotics were used in 786 (93.5％) patients.Among 786 patients,389 (49.5％) were treated empirically with broad-spectrum antibiotics,but only 269 (69.2％) had evidences of bacterial infections.Of the 389 patients with empiric antibiotics use,de-escalation of antibiotics was applied only in 6 (1.54％) patients within 5 days after the initiation of treatment.In 269 patients with evidence of infection,specimen sampling and culture were performed in 248 (92.2％) patients within 3 days,among which 165 samples were positive,and the clinical isolates were mainly multi-drug resistant gram negative bacilli and colonized bacteria in oropharyngeal cavity.De-escalation was applied only in 4 (1.49％,4/269) patients with evidences of bacterial infections.Conclusion Broad-spectrum antibiotics as initial empiric therapy is common for patients in ICU,however de-escalation of empiric therapy is rarely applied even in patients with positive results in pathogen isolation and culture.