The impact of reported beta-lactam allergy in hospitalized patients with hematologic malignancies requiring antibiotics

Abstract

Background. Patients hospitalized with hematologic malignancy are particularly vulnerable to infection. The impact of reported beta-lactam (BL) allergy in this population remains unknown. Methods. This was a retrospective cohort study of adult inpatients with hematologic malignancy admitted at 2 tertiary care hospitals from 2010 through 2015. The primary outcome was hospital length of stay (LOS) after administration of the first antibiotic. Secondary outcomes included readmission, mortality, complications, hospital charges, and antibiotic usage. Our goal was to define the impact of BL-only allergy (BLOA) label on clinical outcomes compared to those with no BL allergy (NBLA) in hematologic malignancy inpatients who required systemic antibiotics. Results. In our cohort (n = 4671), 38.3% had leukemia, 4.9% had Hodgkin lymphoma, 36.1% had non-Hodgkin lymphoma, and 20.7% had multiple myeloma. Among patients, 35.1% reported antibiotic allergy, and 14.1% (n = 660) had BLOA (including 9.3% with penicillin-only allergy and 3.3% cephalosporin-only allergy). Patients with BLOA had longer median LOS compared to patients with NBLA (11.3 vs 7.6 days, P < .001), which remained significant after multivariable adjustment. Patients with BLOA also had significantly worse outcomes in terms of mortality rate at 30 days (7.6% vs 5.3%, P = .017) and 180 days (15.8% vs 12.2%, P = .013), 30-day readmission rate, Clostridium difficile rate, hospital charges ($223 046 vs $173 256, P < .001), antibiotic classes used, and antibiotic duration. Conclusions. In hospitalized patients with hematologic malignancy, patients with reported BL allergy had worse clinical outcomes and higher healthcare cost than those without BL allergy label.