Performance and clinical significance of direct antimicrobial susceptibility testing on urine from hospitalized patients

BACKGROUND: Urinary tract infections (UTIs) are common infections in the community and the hospital. With increasing antimicrobial resistance, specifically in the Gram-negative uropathogens, reliable, rapid antimicrobial susceptibility data would be useful to guide antimicrobial treatment. Direct antimicrobial susceptibility testing (DST) of urine with microscopic evidence of Gram-negative bacterial infection and its clinical significance was investigated in this study.

METHODS: DST was performed by Kirby-Bauer disk diffusion method using undiluted urine as a non-standardized inoculum. Urine specimens with Gram-negative bacteria on microscopy were included. DST results from growth of Gram-negative bacteria were compared to routine antimicrobial susceptibility testing by Phoenix automated system (AST). Errors were scored as 'very major error' if susceptible by DST but resistant by AST and as 'major error' if resistant by DST but susceptible by AST. All other discrepancies were defined as 'minor error'. Discrepancies were resolved by determination of minimum inhibitory concentrations (MICs) using Etests. After discrepancy analysis, errors were scored as above using the Etest as the reference method. For analysis, specimens were divided into 3 categories: category A: 1 isolate found by DST as well as by routine culture; category B: 1 isolate detected by DST, but more than 1 isolate found on routine culture; category C: more than 1 isolate found by both DST and routine culture. The clinical significance of DST was determined prospectively by investigating the potential impact of DST on antimicrobial therapy.

RESULTS: One hundred and sixteen urine specimens were included. For DST and AST there was agreement in 96% of 1152 comparisons in category A (n = 100), 88% of 41 comparisons in category B (n = 4), and 88% of 110 comparisons in category C (n = 12). The 64 discrepancies included 18 very major errors, 7 major errors, and 39 minor errors. Eight very major errors and 11 minor errors were not investigated because the isolates were not available. After Etest MIC determination for the 45 remaining discrepancies, DST showed 1 very major error, 1 major error, and 8 minor errors in category A, none in category B, and 5 major errors and 4 minor errors in category C. Antimicrobial therapy for UTI was prescribed for 53 patients. For 4 patients (8%) therapy was adjusted based on DST because of antimicrobial resistance and for 12 patients (23%) antimicrobial treatment could have been streamlined.

CONCLUSIONS: DST on urine is reliable in monobacterial Gram-negative infections. With increasing antimicrobial resistance, DST can make an important contribution to patient management and reduce the use of broad-spectrum antimicrobials.