2026 Southeastern Residency Conference

Background: Penicillin and sulfonamide allergies are reported by many patients, yet most are not clinically significant. Documented allergies lead to the use of alternative agents such as fluoroquinolones, and this inappropriate use has been linked to increased rates of antimicrobial resistance. While studies have demonstrated the impact antibiotic allergies can have on patient outcomes and healthcare costs, there is minimal literature evaluating the influence documented allergies, whether true or not, may have on antimicrobial susceptibilities. The purpose of this study is to evaluate the relationship between documented antibiotic allergies and antimicrobial susceptibilities for gram-negative bloodstream isolates.

Methods: This is a multicenter, retrospective, cohort study of adult patients admitted to the Prisma Health System with confirmed gram-negative bacteremia from March 1, 2021, to August 1, 2025. The primary outcome is the difference in antibiotic susceptibilities (% susceptible vs % non-susceptible) of isolates between patients with and without documented beta-lactam allergies. Secondary outcomes include the difference in antibiotic susceptibilities among patients with and without fluoroquinolone and sulfonamide allergies and assessing variation in antimicrobial susceptibilities by severity of beta-lactam allergies. Baseline characteristics and outcomes were assessed using descriptive and inferential statistics.

Results: This retrospective cohort study included 500 patients, of which 92 had a beta-lactam allergy present on their chart. The most common bacteria isolated were E. coli (46%), K. pneumoniae (20%), P. mirabilis (10%), and P. aeruginosa (9%). For the primary outcome of rates of non-susceptible isolates in patients with and without a documented beta-lactam allergy, there was no statistically significant difference between the two groups for Enterobacterales or Pseudomonas species for any antimicrobials evaluated. However, patients with a beta-lactam allergy did have numerically higher rates of non-susceptible Enterobacterales isolates for cefazolin, ciprofloxacin,  and sulfamethoxazole-trimethoprim, with a difference of 5%, 5% and 9%, respectively. The secondary outcomes evaluating rates of non-susceptible isolates in patients with and without a fluoroquinolone allergy and with and without a sulfonamide allergy showed no statistically significant difference between groups. Lastly, when stratified by severity of beta-lactam allergy, it was found that patients with a high severity beta-lactam had the highest proportion of non-susceptible Enterobacterales isolates for ampicillin (67%). The medium severity beta-lactam allergy group had the highest proportion of non-susceptible isolates for ceftriaxone (29%), ciprofloxacin (43%), levofloxacin (100%), and sulfamethoxazole-trimethoprim (43%).

Conclusions: Overall, the presence of beta-lactam, fluoroquinolone, or sulfonamide allergies did not impact antibiotic susceptibilities. Almost half of beta-lactam allergies were documented as an “unknown” reaction which emphasizes the improvements that can be made in allergy documentation and reconciliation. This study was limited by the small population of patients with fluoroquinolone or sulfonamide allergies, the subjective manner of allergy assessment, and the process of manual chart review.