Effect of an intensive care unit rotating empiric antibiotic schedule on the development of hospital-acquired infections on the non-intensive care unit ward.

OBJECTIVE: We have previously shown that a rotating empirical antibiotic schedule could reduce infectious mortality in an intensive care unit (ICU). We hypothesized that this intervention would decrease infectious complications in the non-ICU ward to which these patients were transferred.

DESIGN: Prospective cohort study.

SETTING: An ICU and the ward to which the ICU patients were transferred at a university medical center.

PATIENTS: All patients treated on the general, transplant, or trauma surgery services who developed hospital-acquired infection while on the non-ICU wards.

INTERVENTIONS: A 2-yr study consisting of 1-yr non-protocol-driven antibiotic use and 1-yr quarterly rotating empirical antibiotic assignment for patients treated in the ICU from which a portion of the patients were transferred.

MEASUREMENTS AND MAIN RESULTS: There were 2,088 admissions to the non-ICU wards during the nonrotation year and 2,183 during the ICU rotation year. Of these patients, 407 hospital-acquired infections were treated during the nonrotation year and 213 were treated during the ICU rotation (19.7 vs. 9.8 infections/100 admissions, p <.0001). During the ICU rotation year a decrease in the rate of resistant Gram-positive and resistant Gram-negative infections on the non-ICU wards occurred (2.5 vs. 1.6 infections/100 admissions, p =.04; 1.0 vs. 0.4 infections/100 admissions, p =.03). Subgroup analysis revealed that the decrease in resistant infections on the wards was due to a reduction in resistant Gram-positive and resistant Gram-negative infections among non-ICU ward patients admitted initially from areas other than the ICU implementing the antibiotic rotation (e.g., home, other ward, or a different ICU) (1.8 vs. 0.5 infections/100 admissions, p =.0001; 0.7 vs. 0.2 infections/100 admissions, p =.02), not due to differences for those transferred to the ward from the rotation ICU (10.4 vs. 9.7 infections/100 admissions, p = 1.0; 4.3 vs. 1.9 infections/100 admissions, p =.3). No differences in infection-related mortality were detected.

CONCLUSIONS: An effective rotating empirical antibiotic schedule in an ICU is associated with a reduction in infectious morbidity (hospital-acquired and resistant hospital-acquired infection rates) on the non-ICU wards to which patients are transferred.