Variation in Reporting of Penicillin Allergy and its Consequences: an Evaluation of 13 Hospitals

BACKGROUND: Penicillin allergy is the most common antibiotic allergy noted within medical records, and its inaccurate reporting leads to increased use of alternative antibiotics that may be less effective, broader in spectrum, more toxic, and costly. METHODS: We retrospectively reviewed the reported allergies to penicillin in patients cared for at 13 hospitals within one health system over a 3 month period (June-August 2016). The data were abstracted from the electronic medical records on penicillin allergy status for both inpatient and outpatient visits. Hospitals were compared on their use of systemic antibiotics for inpatients. The proportions of total defined daily doses (DDD) for quinolones, aztreonam, carbapenems, cephalosporins, and penicillins were compared. Spearman’s rank and Pearson’s correlation were used to evaluate the strength of the relation between increased penicillin allergy reported and the use of the different antibiotic classes. RESULTS: 23,290 of 169,912 (13.7%; range 8%–20%) patients from 13 hospitals were reported penicillin allergic. There was a strong correlation between the proportion of patients with penicillin allergy and quinolone use (rho=0.77; P = 0.002; Figure 1), cephalosporins excluding fourth-generation (r=0.70; P = 0.007; Figure 2), and a weaker correlation with carbapenem use (rho=0.52; P = 0.168) and aztreonam (r=0.53; P = 0.06). On the other hand, penicillins had a moderate negative correlation (r=-0.58; P = 0.036; Figure 3), and extended spectrum penicillins had a strong negative correlation (r=-0.72; P < 0.005). Fourth-generation cephalosporin use did not correlate with the penicillin allergy rate (rho=-0.03; P = 0.92). CONCLUSION: Reported penicillin allergy varies between hospitals and higher reported allergy is associated with more quinolone and cephalosporin use, and less use of penicillin-based regimens. Adequate documentation of penicillin allergy may promote the choice of more optimal regimens when treating patients. DISCLOSURES: All authors: No reported disclosures.