P2.178: A Characterization of Serum Sickness-Like Reaction Diagnosis in the Pediatric Emergency Department

Background: The delayed reaction ‘serum sickness-like reaction’ (SSLR) commonly occurs in children. It is characterized by combinations of fever, rash, and arthralgia and is often attributed to antibiotics. Although SSLR has been viewed as high-risk and providers have been reluctant to re-expose children to the “culprit” antibiotic, emerging literature supports that SSLR temporarily associated with antibiotics is more likely to be a hyperinflammatory response during a viral infection. Supporting this, recent literature where children have been assessed in specialty allergist clinics has suggested children can be delabeled of the SSLR culprit antibiotic in the majority of cases. We investigated the characterization and interventions for children diagnosed with SSLR in a pediatric emergency department (PED) where most children with SSLR are seen.

Methods: Children evaluated in the Children’s Wisconsin PED with a discharge diagnosis of serum sickness or SSLR between 2015-2024 were included. Presenting symptoms, suspected antibiotic causing reaction, recommended treatments, and subsequent allergy labeling were reviewed and entered into REDCap summarizing variables using counts and proportions.

Results: 241/294 (82%) reviewed encounters cited an antibiotic as the SSLR culprit. “Culprit” antibiotics included amoxicillin (153 [52%]), cefdinir (35 [12%]), and amoxicillin-clavulanic acid (32, [11%]). However, another 32 cases (11%) were attributed to a viral illness, and 18 (6%) had no attributable cause. 148 (50%) patients had symptom onset in an 8–14-day window while 80 (27%) had onset in 4-7 days. 92 (31%) patients reported rash, edema, and joint pain/swelling, 52 (18%) reported rash and edema, and 42 (14%) reported rash alone. Discharge recommendations included ibuprofen (198 [67%]), diphenhydramine (172 [59%]), and cetirizine (97 [33%]). Only 71 (24%) had laboratory studies completed, and only one patient had a respiratory pathogen panel collected, which was negative. As a result of their diagnosis, a vast majority of patients had an antibiotic allergy label added in the EHR at the encounter and very few had subsequent allergy follow-up or delabeling.

Conclusion: Our retrospective review suggests that features leading to a diagnosis of SSLR are inconsistent and, in many where an antibiotic was cited, rash was the only symptom. The lack of work-up for an underlying viral cause and subsequent antibiotic labeling in the EHR suggest SSLR may be over-diagnosed, and lead to decreased future use of appropriate first-line antibiotics. Education of pediatricians and future studies are needed to prevent the negative impact of true and false diagnosis of SSLR on public health and antibiotic stewardship.