Background: Rapid diagnostic tests (RDTs) are becoming increasingly employed to assist in the management of gram-negative bacteremia. Given the diversity of pathogenic organisms and resistance mechanisms, clinical data regarding optimal management using RDT is lacking. Moreover the choice of optimal RDT platform remains elusive, as comparisons are limited to sensitivity and specificity in small samples. This study compared a key clinical outcome, potential antimicrobial decisions, based on results of different commonly used RDT platforms using a novel methodology termed Desirability of Outcome Ranking Management of Antimicrobial Therapy (DOOR-MAT).
Materials/methods: Retrospective observational study at University of Maryland Medical Center from 08/2018 – 11/2019 of adult patients with gram-negative bacteremia comparing Verigene® Blood Culture (VBC) to BIOFIRE® FILMARRAY® Blood Culture Identification (BCID) and BCID2 research use only (RUO) panels for clinical blood cultures. Verigene® was part of standard of care, BCID and BCID2 were run on discarded frozen samples. The RDT results and local susceptibility data were applied by an Infectious Diseases-trained pharmacists to make decisions regarding potential antimicrobial selection. DOOR-MAT, a partial credit scoring system, was used to compare antimicrobial decisions as a function of final phenotypic susceptibly patterns as determined by VITEK® 2 automated susceptibility testing (Figure 1). DOOR-MAT scores were compared between panels using Kruskal-Wallis with p < 0.05 statisticaly significant.
Results: A total of 103 patients with positive clinical cultures for gram-negative bacteria were included. The average DOOR-MAT score for VBC was 85.8 (SD 25.7) and median score was 100 (IQR 62.5, 100). BCID resulted in an average score of 60.8 (SD 33.4) and median 50 (IQR 50,100). BCID2 (RUO) demonstrated an average score of 89.7 (SD 24.7) and median score 100 (IQR 100, 100). Overall, BIOFIRE® FILMARRAY® BCID2 (RUO) produced the highest scores for optimal therapy. There was a significant difference in DOOR-MAT scores (p < 0.0001) between tested panels.
Conclusions:. The BIOFIRE® FILMARRAY® BCID2 (RUO) performed best among the three panels tested. Use of a partial credit scoring system such as the DOOR-MAT allows for comparisons between RDT systems beyond sensitivity and specificity allowing for enhanced clinical interpretation.