Labelling patients as allergic to beta-lactam antibiotics is associated with periprosthetic joint infection up to five years following knee arthroplasty

Abstract

Aims: Periprosthetic joint infections (PJIs) represent a significant complication of total knee arthroplasty (TKAs). However, the influence of drug or beta-lactam (BL) antibiotic allergy labelling of patients on PJI remains largely unknown. In this study, we examine the association between patients labelled with a BL allergy and the occurrence of PJI among patients undergoing TKA. We also assess the prevalence of incorrect patient labelling and explore the feasibility of a multidisciplinary drug allergy testing initiative to detect mislabelling. Methods: Longitudinal data from all patients who underwent TKA between January 1993 and December 2021 were analyzed. We investigated the association between different risk factors and PJI, with particular focus on patients labelled as having an antibiotic drug allergy. The outcomes of patients with and without a labelled BL allergy were compared. Additionally, patients labelled as having a BL allergy and who had undergone or were scheduled for TKA were prospectively investigated by formal allergy assessment. Results: Out of 4,730 TKAs, the overall incidence of PJI was 1.0% (47/4,730). Patients labelled as having a BL allergy had a higher incidence of PJI within the first five years post-TKA compared to those without (3.0% (5/165) vs 0.7% (34/4,565); p = 0.001). The presence of a BL allergy label was identified as an independent risk factor for PJI (hazard ratio 4.86 (95% CI 2.05 to 11.53); p < 0.001). Following negative drug provocation testing, the majority of patients (95% (21/22)) evaluated with BL allergy labels were successfully delabelled. Conclusion: In this longitudinal study, patients labelled as having a BL allergy were associated with having increased risk of PJI following TKA, particularly within the first five years. Given the high rate of patients being mislabelled, we recommend that patients labelled as having a BL allergy should be prioritized for formal allergy assessment and evaluation. Further studies on the impact of preoperative antibiotic allergy delabelling initiatives should be encouraged.