Background
The optimal duration of therapy for bloodstream infections due to Gram-negative bacteremia has been poorly defined. The objective of this study was to determine if short courses of antimicrobial therapy were noninferior to intermediate and long courses of antimicrobial therapy for Gram-negative bacteremia.
Methods
We reviewed medical records on patients with a documented Gram-negative bacteremia from August 2006 to November 2013. Patients meeting eligibility criteria were placed in the short-course, intermediate-course, and long-course treatment groups if their duration of antimicrobial therapy was less than or equal to 7 days, between 8 to 14 days, or greater than 14 days, respectively. The primary outcome was the achievement of a clinical response at the end of therapy. Secondary outcomes included achievement of microbiological cure at the end of therapy, and association of clinical response in each group with the causative organism, source of bacteremia, and time to defervescence.
Results
Of 406 cases of Gram-negative bacteremia, 178 cases met eligibility criteria. Median age was 64 years with 67% females. Median SAPS II (Simplified Acute Physiology Score II) was 33 points. The most common infecting pathogen was Escherichia coli (46%), followed by Klebsiella pneumoniae (22%). The most common source of bacteremia was the urinary tract (53%), followed by indwelling catheters (14%). Clinical response rates at the end of therapy were 78.6%, 89.0%, and 80.6% for the short-course, intermediate-course, and long-course treatment groups, respectively (P = 0.2). Microbiological cure rates at the end of therapy were 83.3%, 89.0%, and 91.7% for the short-course, intermediate-course, and long-course treatment groups, respectively (P = 0.7). Logistic regression analysis did not reveal any association of clinical response in each group with infecting organism (P = 0.4), source of bacteremia (P = 0.5), or time to defervescence (P = 0.6).
Conclusions
Short-course therapy for Gram-negative bacteremia appears to achieve similar clinical response rates and microbiological cure rates compared with intermediate- and long-course therapy.
