Prevalence and Risks of Discordant Antibiotic Therapy for Bloodstream Infections

Roughly 1 in 5 patients with bloodstream infections in US hospitals received discordant empirical antibiotic therapy, according to results published in The Lancet Infectious Diseases. The receipt of discordant therapy, defined as receiving antibiotics on the day blood culture samples were drawn, was closely associated with antibiotic-resistant pathogen infections and increased odds of overall mortality.

Investigators aimed to establish the population-level burden, predictors, and mortality risk of in-vitro susceptibility-discordant empirical antibiotic therapy in patients with bloodstream infections as the prevalence and effects to this point are unclear. Investigators conducted a retrospective cohort analysis of electronic health record data from 131 hospitals in the United States between January 1, 2005 and December 31, 2014. Patients with suspected-and subsequently confirmed-bloodstream infections treated empirically with systemic antibiotics were included.

Investigators identified 21,608 patients with bloodstream infections who received empirical antibiotic therapy on the day of first blood culture collection; 4165 (19%) received discordant empirical antibiotic therapy. Receiving discordant therapy was independently associated with increased risk of mortality (aOR, 1.46; 95% CI, 1.28-1.66; P <.0001). This relationship was unaffected by presence or absence of resistance of sepsis or septic shock. Receiving discordant therapy was a strong predictor of infection with antibiotic-resistant species. Most incidences of discordant therapy and associated deaths occurred in infections caused by Staphylococcus aureus or Enterobacterales.

Despite the study population being drawn from a well-distributed cohort of US hospitals, the findings might not be generalizable to other hospitals in the United States or regions worldwide. Investigators were unable to assess the adequacy of source control given the absence of clinical documentation. Exclusions of patients with polymicrobial bloodstream infections ─ due to the difficulty merging microbiology and susceptibility datasets for multi-organism samples ─ could have affected estimates.

They added, “Residual confounding might have persisted despite our efforts at risk adjustment.” While the use of imputed missing susceptibilities based on consensus adjudication maximized the available data, it could have introduced some element of subjectivity. Inpatient encounters with bloodstream infection for which no antibiotic order was identifiable in the database might have eliminated some cases where administration of antibiotics was truly missed. Due to the data being insufficiently granular, investigators were unable to generate hourly estimates of associations between the time to antibiotic administration and outcomes.

Investigators concluded that this retrospective population-level analysis emphasizes the importance of early identification of bloodstream pathogens and their resistance profiles, especially for S. aureus and Enterobacterales. They believe this early identification is “a means to potentially mitigate the mortality burden attributable to discordant empirical antibiotic therapy.” They further recommended that researchers around the globe with similar data sets perform similar assessments of patterns and effects of empirical antibiotic prescribing as this could improve local awareness and identify areas for improvement.  

Disclosure: One study author declared affiliations with the pharmaceutical industry. Please see the original reference for a full list of author’s disclosures.


Kadri SS, Lai YL, Warner S, et al. Inappropriate empirical antibiotic therapy for bloodstream infections based on discordant in-vitro susceptibilities: a retrospective cohort analysis of prevalence, predictors, and mortality risk in US hospitals. Published online September 8, 2020. Lancet Infect Dis. doi: 10.1016/S1473-3099(20)30477-1