D1.219 - Impact of a Penicillin Allergy Delabeling Program on Antibiotic Selection and Perioperative Outcomes: A Prospective Comparative Cohort Study

Mislabeling patients with penicillin allergies often leads to suboptimal antibiotic perioperative use, reducing the effectiveness of prophylaxis and increasing post-operative complications. Our objective was to evaluate the impact of a Penicillin Allergy DeLabeling (PADL) program on perioperative care by comparing intraoperative antibiotic selection and post-operative infection rates.

From January 1 to December 31, 2017, a prospective study was conducted at the McGill University Health Center’s two sites. Both sites shared surgical and anesthesiology departments, but only one site implemented a PADL program where patients flagged as penicillin-allergic were either de-labeled or maintained their allergy status. Antibiotic prophylaxis usage and post-operative outcomes were compared.

At the first site where penicillin allergy de-labeling (PADL) was implemented, 5.3% of 10,666 patients were flagged as allergic. Of the 116 de-labeled patients, 72.4% received cefazolin, and only 3.4% required vancomycin. Among the 447 non-de-labeled patients, cefazolin use dropped to 22.6%, while 21.7% required vancomycin. In contrast, at the other site without PADL, 7.8% of 9,686 patients were labeled as allergic. Only 6.3% received cefazolin, while 24.3% required vancomycin. Post-operative complication rates were significantly lower among de-labeled patients (3%, 95% CI: 0.79%–9.32%) compared to those retaining allergy labels (8.7%, 95% CI: 6.69%–10.7%). De-labeled patients experienced minimal complications, including single cases of wound infection, urinary tract infection, and blepharitis. In contrast, non-de-labeled patients faced a higher frequency of severe complications, including wound infections (n=18), pneumonia (n=8), cellulitis (n=6), sepsis (n=5), and miscellaneous conditions (n=17).

Similar rates of suspected penicillin allergy were noted at both sites, with no differences in peri-operative antibiotic use among non-de-labeled patients. De-labeled patients had higher use of prophylactic antibiotics (84.5% vs. 59.9%) and cefazolin as the first-line antibiotic (72.4% vs. 12.6%), with reduced reliance on vancomycin or clindamycin (7.7% vs. 41.4%). Vancomycin use decreased by 40%, irrespective of MRSA status. Post-operative complications were also lower among de-labeled patients (3% vs. 8.7%, absolute reduction 5.7%; 95% CI: 1.1%, 10.2%). These results highlight the benefits of penicillin allergy de-labeling in optimizing antibiotic use, reducing complications and highlighting the critical role of accurate allergy documentation and re-evaluation in improving surgical outcomes.