Main Menu Browse by Day or Program Author Keywords

86 Real-time PCR Versus Traditional Culture Methods for the Identification of Staphylococcal Bacteremia and Its Effect on Antibiotic Administration

Friday, March 19, 2010
Grand Hall (Hyatt Regency Atlanta)
Christopher J. Bettacchi, MD , University of Alabama at Birmingham, Birmingham VA Medical Center, Birmingham, AL
Kinley Beck , University of Alabama School of Medicine, Birmingham, AL
John W. Baddley, MD , University of Alabama at Birmingham, Birmingham VA Medical Center, Birmingham, AL
Ken B. Waites, MD , University of Alabama at Birmingham, Birmingham VA Medical Center, Birmingham, AL
Mukesh Patel, MD , University of Alabama at Birmingham, Birmingham VA Medical Center, Birmingham, AL
Background: Speciation of staphylococci and reporting of antibiotic susceptibilities can take 3-5 days with traditional blood culture methods.   Polymerase chain reaction (PCR) amplification techniques may allow for more rapid identification of samples; however, use of rapid molecular tests to identify staphylococci in blood and the impact  on clinical outcomes needs exploration.

Objective: To assess the use of a rapid PCR-based diagnostic test to identify staphylococci in blood cultures and the impact on  antimicrobial prescribing.

Methods: We conducted a quasi-experimental cohort study of patients at the Birmingham Veterans Affairs Medical Center. Unique patients (pts) with at least one blood culture growing gram-positive cocci in clusters (GPCC) were included.  Between  Jan 2008 and Feb 2009  pathogen identification was performed using traditional culture techniques, and between  March 2009 & Nov 2009 pathogen identification using real-time PCR ( Cepheid GeneXpert® System) was performed. Medical records were reviewed to obtain clinical data.  All cases were reviewed by 2 infectious diseases physicians to determine appropriateness of antibiotic (abx) selection at 12 hours after culture growth.  Comparisons of culture and PCR groups were made using Chi square and T tests.

Results: 107 patients met study criteria: 75 in the culture group, 32 in the PCR group. Mean age was 62.7 years, 98% were male, 53% Caucasian, and mean Charlson Comorbidity Index was 5.  Demographics were similar in the two groups.  Common comorbidities included diabetes mellitus (46%), coronary artery disease (27%), chronic kidney disease (38%), dialysis (25%), and malignancy (22%). Mean length of stay in days (d) after culture was not significantly different between the culture (12.1d) and PCR tested (8.0d) patients, P=0.41.  Bacteria identified included MRSA (39%), MSSA (21%), coagulase-negative staphylococci (37%), and other(3%).  Mean time to culture growth was 0.99d and similar for both groups.  Time from culture growth to final organism identification (ID) was shorter in the PCR group (0.24d vs. 2.16d; P < 0.0001), as was  time from  date of culture collection to final organism ID (1.00d vs. 3.18; P < 0.0001).  Timing of initial abx therapy after culture collection was similar in both groups (mean 0.3d).  Receipt of appropriate abx therapy within 12 hours of culture growth was more frequent in the PCR group (72% vs. 39%; P = 0.003). Empiric abx therapy was started in 100% of patients in the culture group and 78% of patients in the PCR group (P <0.001).

Conclusions: Real-time PCR for blood culture identification was more rapid than traditional blood culture methods and was associated with earlier appropriate antimicrobial therapy.   Impact on patient outcomes and use in antimicrobial stewardship requires further study.