Socioeconomic and Demographic Differences in Rates of Penicillin Allergy Labels in Children

Abstract Corresponding author: Rana E. El Feghaly MD, MSCI, Department of Pediatrics, Division of Infectious Diseases, Children’s Mercy Kansas City, 2401 Gillham Rd, Kansas City, MO 64108. [email protected]; 816 234 3061 Conflict of Interest: REE and AN received funding from Merck for an investigator-initiated grant on reducing health inequity in antimicrobials prescribing for respiratory infections in pediatric urgent cares. All other authors report no conflict of interest. Background Penicillins are the most widely prescribed antibiotics. Of the 10% of children with a penicillin allergy label (PAL), 95% do not exhibit a type 1 hypersensitivity allergic reaction following exposure. This study aimed to investigate whether sociodemographic factors, such as race, ethnicity, primary language spoken, insurance type, and area deprivation index (ADI), a tool used to quantify relative socioeconomic conditions of neighborhoods, are associated with PALs in children. Methods We performed a retrospective chart review of children ≤19 years evaluated in our pediatric urgent care centers and emergency departments from July 2022-June 2023. We collected patients’ self-reported race, ethnicity, language, use of interpreter, and insurance type, in addition to zip code used to calculate ADI. Multivariable logistic regression models identified groups with higher odds of PALs. Results We reviewed 183,692 patients’ charts and found 11,669 (6.3%) patients with PAL. After adjusting for different sociodemographic variables, compared to White Non-Hispanic (NH) children, children of almost all races and ethnicities were less likely to have PALs. Compared to children with Medicaid, children with commercial insurance were more likely to carry a PAL (aOR 1.11, CI 1.03, 1.2, p 0.007), while patients who self-pay were less likely to have PALs (aOR 0.67, CI 0.56, 0.8, p <0.001). Compared to patients not requiring language services, patients who required Spanish language services were less likely to have PAL s(aOR 0.57, CI 0.45, 0.72, p <0.001). Finally, patients from higher ADI deciles (i.e., the neighborhoods with greater socioeconomic disadvantages) were less likely to have PALs (Tables 1 and 2). Conclusions In addition to race, socioeconomic status, reflected by insurance and ADI, was also associated with differences in PAL rates. We are now conducting a quality improvement project to reduce inaccurate PALs, with an increased awareness of socioeconomic and demographic variables, to improve health equity among our patients. Table 1: Comparison of the demographics of children with penicillin allergy labels compared to those without a penicillin allergy label Table 2: Logistic Regression models depicting the odds ratio of a penicillin allergy label in our patient population