Prior Antibiotic Allergies and Adverse Reactions: Results from a Multicenter Cross-Sectional Study


Weisenberg SA, Braykov NB, Schweizer ML, Morgan DJ, Johannsson B, Kelesidis T, Uslan DZ, Young H, Cantey JB, Perencevich EN, Srinivasan A, Septimus EJ, Laxminarayan R. Prior Antibiotic Allergies and Adverse Reactions: Results from a Multicenter Cross-Sectional Study (Poster #769). IDWeek. October 2012. San Diego, CA. 




Adverse reactions (AR) to antibiotics are common among outpatients, accounting for ~20% of drug-related AR visits to emergency rooms. Little is known about the prevalence of ARs among inpatients. This retrospective study describes patient-reported antibiotic allergies on admission, their effect on outcome, and ARs associated with antibiotic use in 6 acute care facilities of varying size and location.


Cross-sectional study using retrospective chart review of 1,200 adult inpatients (pts), hospitalized (>24hrs) in Veterans Affairs (n=1), teaching (n =2), non-teaching (n=3) hospitals, and receiving ≥1 antibiotic doses on 4 index dates corresponding to a random hospital day and chosen at equal intervals through a 1-year study period (9/2009-10/2010). Medical records were reviewed for antibiotic allergies on admission, demographics, comorbidities, diagnostics, duration and indication for prescribed antibiotics, treatment outcomes and ARs. Therapy failure was defined as lack of treatment response based on clinical evidence available in record. Proportions were compared using Pearson’s χ2 test. 


History of antibiotic allergies was present in 311/1,200pts (25.9%) that reported 429 allergies (83 to >1 antibiotic class). The most common allergies were to penicillins (185), sulfonamides (99) and cephalosporins (35). Rash was the most common prior adverse reaction (21% n=90), but 46% of prior adverse reactions were unknown (n= 196). One hundred seven pts (9%) received an antibiotic to which they had a prior documented allergy. Pts with prior allergies had higher odds for therapy failure (OR = 1.4, p = 0.08) than non-allergic patients despite lower Charlson scores (p = 0.002) and a lower likelihood of immunosuppresion (OR=0.56, p = 0.01). New ARs were noted for 147/1,200 study pts (12%), including diarrhea (8%), nausea (1.3%), hepatic or renal failure (1.3%), and rash (0.6%). Patients with new ARs had increased odds of therapy failure (OR 1.68, p = 0.02). 


Patient-reported antibiotic allergies in inpatients are common and may limit optimal treatment, though specific allergies are often unconfirmed. New adverse reactions occur in 12% of inpatients and are associated with treatment failure.  Improved documentation of antibiotic allergies could minimize risk of treatment failure and ARs.