Abstract
Background
Hospitalized pediatric patients with bacteremia receive broad-spectrum antibiotics while organism identification and antibiotic susceptibilities are pending. Direct susceptibility testing (DST) using unstandardized Kirby-Bauer disk diffusion provides early results before standardized (final) antimicrobial susceptibility testing (AST) is available. The accuracy of DST in comparison with AST has been poorly studied. If DST is highly accurate, it could facilitate earlier de-escalation of antibiotics.
Methods
Retrospective cohort study of all positive blood cultures at Boston Children’s Hospital between January 1, 2017 and October 20, 2019. Isolates were eligible for inclusion if a DST result was available. Isolates were excluded if more than one organism grew from a blood culture or if a patient had repeat blood cultures positive for the same organism within 14 days. Patient characteristics and antibiotic orders were identified via a local data warehouse. Positive and negative predictive value (PPV: accuracy of susceptibility on DST in identifying susceptibility on AST result; NPV: accuracy of DST in identifying non-susceptibility on AST result) and 95% confidence interval were calculated for each bug-drug combination. Antibiotic Spectrum Index was evaluated at 3 time points to assess change in antibiotic prescribing after availability of DST and AST results.
Results
496 patients (median age: 51 months, IQR: 7–165 months) with 603 positive blood cultures were included in the final analysis. PPV of DST was ≥96%for most organism-antibiotic pairs (Table). NPV of DST varied substantially across organism-antibiotic pairs and was frequently lower than PPV. The proportion of patients with more narrow spectrum antibiotic orders increased after the DST result and again after the AST result (Figure).
Table
Test performance of direct susceptibility testing (DST) of clinical isolates from patients with bacteremia to identify organism susceptibility to the noted antibiotic. Positive predictive value measures the ability of a susceptible DST result to predict a susceptible AST result. Negative predict value measures the ability of an intermediate or resistant DST result to predict an intermediate or resistant AST result. *Includes Staphylococcus capitis, Staphylococcus epidermidis, Staphylococcus haemolyticus, Staphylococcus hominis, Staphylococcus pettenkoferi, and Staphylococcus warneri. ‡Includes Citrobacter freundii complex, Citrobacter koseri, Enterobacter aerogenes, Enterobacter cloacae complex, Escherichia coli, Escherichia vulneris, Klebsiella oxytoca, Klebseilla pneumoniae, Klebsiella variicola, Pantoea species, Proteus mirabilis, Salmonella species, Salmonella typhi, Serratia liquefaciens, and Serratia marcescens.
Figure
Distribution of spectrum of antibiotics ordered to treat bacteremia in pediatric patients. Antibiotic spectrum, as measured by the antibiotic spectrum index, is represented in this Sankey diagram on the Y-axis. The height of each group indicates the proportion of patients with an aggregate antibiotic spectrum falling in that category at three time points. The ASI of empiric antibiotics was measured just before the DST result to provide time for a treating physician to settle on an empiric regimen. The ASI with the DST result was measured just before the ASI result. The ASI with the AST result was measured 24 hours after the AST result was available. Green bars indicate the proportion of patients within a group whose ASI remained unchanged between time points. Blue bars indicate the proportion of patients within a group whose ASI decreased, and red bars indicate the proportion of patients whose ASI increased between time points.
Conclusion
DST is highly accurate at identifying susceptibility to antibiotics for many bug-drug combinations in pediatric blood culture isolates, but its ability to identify non-susceptibility is less robust. The observed spectrum of prescribed antibiotics was narrower after DST results, suggesting some clinicians may be using the result to de-escalate therapy. DST may be a useful low-cost tool for antimicrobial stewardship.
Disclosures
All Authors: No reported disclosures
Novel electronic biosensor for automated inoculum preparation to accelerate antimicrobial susceptibility testing
