Background: Conceptually, appropriateness of antibiotic therapy includes choice of agent relative
to susceptibility of pathogens as well as dosing, timing of onset, and duration of therapy,
but is most commonly considered in terms of choice of antibiotic. It has been suggested
that inappropriate antibiotic selection can result in increased mortality. This study was performed
to elucidate the role of scheduled, rotating antibiotic therapy in defining mortality
among febrile, infected surgical ICU patients.
Methods: Prospective inception-cohort study of 356 patients during their initial episode of
fever (temperature > 38.2° C), caused by infection diagnosed by positive cultures or direct inspection
(some cases of peritonitis). Collected data included age, gender, admission APACHE
III score, peak temperature, microbial isolates and susceptibility, source of infection, multiple
organ dysfunction score, mortality, and several time intervals (time that cultures were collected,
time from collection to antibiotic prescription, time from collection to antibiotic administration,
duration of therapy).
Results: The mean age was 63 ± 1 years, the mean APACHE III score was 74 ± 2 points, the
mean multiple organ dysfunction score was 8 ± 1 points, and overall mortality was 31%. Neither
the source of infection nor the specific isolate influenced mortality. Antibiotic therapy
was appropriate (covered the isolates) in 94% of cases, and did not influence mortality. Duration
of therapy was identical between groups (5.1 ± 0.3 vs. 5.4 ± 0.3 days, p = 0.61). By logistic
regression (dependent variable = mortality), APACHE III score OR 1.025, 95% C.I.
1.021–1.04) and delayed antibiotic administration (30-min intervals, OR 1.021, 95% C.I.
1.003–1.038) were independent predictors of mortality.
Conclusions: The use of scheduled monthly antibiotic cycling in the surgical ICU is associated
with a high rate of "appropriate" antibiotic therapy, and appears to maintain or improve
resistance patterns. Because antibiotic therapy was mostly appropriate for isolates, initial
inappropriate therapy could not be identified as a risk factor for mortality. However, in
the setting of appropriate antibiotic choice, the prompt initial administration of antibiotics
appears to be crucial for survival, but neither site of infection nor specific pathogen are influential.
