Direct identification and determination of antimicrobial susceptibility of gram negative bacilli using the PhoenixTM FX system in blood cultures flagging positive

Abstract

Bloodstream infections are one of the main causes of morbidity and mortality worldwide. Shortening the turnover time of microbiological analysis leads to a significant reduction of patient morbidity, mortality and medical care expenses. This study aimed to evaluate the performance of Phoenix 100 Instrument (Becton Dickinson, Sparks, MD, USA) system in bacterial identification and the evaluation of antibiotic susceptibility directly taken from positive blood culture bottles in BD BACTEC™ FX (Becton Dickinson, USA) system in patients with the suspicion of bacterial sepsis. In this study, blood culture bottles with a positive signal in BD BACTEC™ FX (Becton Dickinson, USA) system, and in which gram negative bacilli or coccobacilli was observed in Gram staining were evaluated. In our study, direct inoculation to the Phoenix panel from the blood culture bottles giving positive signal was defined as "direct Phoenix method". The blood culture bottles which were taken from hospitalized patients with a suspicion of sepsis, were analyzed comparatively by using BD Phoenix™ (Becton Dickinson, USA) automated microbiology system and "Bruker Biotyper matrix-associated laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS)" (Bruker Daltonics, Germany) in the bottles of patients with the positive signals between August 10, 2015, and December 05, 2015. The effect of the "direct Phoenix method" on the duration of the test and the harmony with conventional methods was demonstrated. A total of 122 gram negative bacteria comprising 95 bacteria in Enterobacteriaceae family, 26 non-fermenting gram negative bacteria (NFGNB) and one Aeromonas spp. were included in the study. In our study, the reporting time of the identification of gram negative bacteria after inoculation with direct Phoenix method ranged from 2.25 hours to 7.84 hours (mean 2.56 hours). When the identification of 122 gram negative bacteria by direct Phoenix method was compared with the identification of Maldi Biotyper (Bruker Daltonics, Germany), an agreement was detected in 96.7% (118/122) of the samples at the genus level and in 86.0% (105/122) of the samples at the species level. Ninety-five bacteria belonging to Enterobacteriaceae family demonstrated an agreement of 95.7% (91/95) and 93.6% (89/95) at the genus and species levels with Maldi Biotyper respectively. Twenty-five NFGNB had an agreement of 96.1% (25/26) and 61.5% (16/26) at genus and species levels with Maldi Biotyper respectively. In our study, the reporting time of antibiotic susceptibility test results of gram negative rods after direct Phoenix method ranged from 7.42 hours to 15.85 hours (mean 12.9 hours). In our study, 2.159 antimicrobial agents were tested for 120 gram negative bacteria (95 strains belonging to Enterobacteriaceae family and 25 NFGNB). Minor error rate was found to be 2.0% with the direct Phoenix method, 1.1% major error rates, and 1.2% very major error rates; making a total of 4.4% (96/2.159). Since error rates in all categories < 10%, very major error rate was < 1.5% and major error rate was < 3%, the direct Phoenix method had an acceptable agreement with the conventional Phoenix method. The direct inoculation of the Phoenix system with culture suspension obtained from positive blood culture bottles decreased reporting time of the identification and determination of the antibiotic susceptibilities for gram negative rods that cause bacteraemia. This result could be important in the clinical benefits for the patients as well as financial savings for the hospital.