Methods: We conducted a quasi-experimental, non-equivalent, pre- and post-intervention study between 1/2012 and 8/2013. Retrospective chart review was performed on 500 patients (250 pre-intervention (Group 1) and 250 post-intervention (Group 2). Patients < 18 years of age, outpatients, patients who were discharged or who died within 24 hours of blood culture (BC) results, and those with BC that yielded Gram-negative, anaerobic, or mixed organisms were excluded. Clinical characteristics, APACHE II score, time to effective therapy (ET) (antimicrobials that would cover the pathogen) and optimal therapy (OT) (antimicrobials that would most narrowly cover the pathogen), length of ICU and hospital stay (LOS), antimicrobial cost, and mortality data were collected. Group 1 BC isolates underwent standard microbiologic identification which included a real-time PCR for fem and mecA on staphylococcal isolates. Group 2 BC isolates identified by Gram stain as Gram-positive cocci during the day shift, 7 days per week, were tested with VBC-GP. Pharmacists participated in AS efforts in both groups (notified MDs with results). Group 1 and Group 2 were statistically matched using propensity scoring. Group 1 and Group 2 were statistically matched using propensity scoring.
Results: 386 patients met inclusion criteria and underwent propensity matching based on age, diabetes, and malignancy. 120 matched pairs were used to compare outcomes of interest. The implementation of VBC-GP was associated with a mean reduction in the time from BC collection to receipt of OT of 10 hours when compared to existing identification methods (2.21 vs. 1.79 days, p<0.02). No significant differences were found among other outcomes of interest including time from blood culture collection to ET, LOS, antibiotic cost, or mortality.
Conclusion: The use of VGP-BC was associated with a modest improvement in AS even in a setting where advanced molecular testing for staphylococcal BC isolates was already in place. For the majority of microbiology labs without in-house molecular diagnostics, the use of this rapid test may have even larger benefits in terms of reduction in antimicrobial use and time to OT.
S. Diederich, None
R. Thomson Jr., None
M. Suseno, None