Background: Eosinophilia in pediatric patients encompasses a broad spectrum of etiologies, ranging from allergic disorders to hematologic malignancies. Understanding the distribution and predictors of eosinophilia is crucial for accurate diagnosis and management. Often these patients are referred to several subspecialties including allergy/immunology and hematology leading to confusion on how best to care for these patients.

Objectives: This study aimed to determine the frequency and distribution of eosinophilia causes in pediatric patients, explore demographic and clinical predictors, and assess the relationship between eosinophilia severity and etiology.

Methods: A retrospective chart review was conducted at Children's National Hospital, encompassing cases from January 2015 to December 2023. This is a project undertaken as a Quality Improvement Initiative and does not constitute human subjects research. As such it was not under the oversight of the Institutional Review Board. Inclusion criteria were patients aged 0–21 years with an absolute eosinophil count (AEC) > 0.3 × 10⁹/L and at least one follow-up or clinical encounter. Data sources included electronic medical records, laboratory information systems, and clinical documentation. Variables extracted encompassed demographics, clinical data, exposure history, laboratory data, and diagnoses. Multinomial logistic regression was utilized to assess the effect of patient characteristics on eosinophilia causes, with “Allergy only” as the reference category. Logistic regression analyses were also performed to evaluate predictors for asthma, eczema, allergic rhinitis, and food allergy.

Results: Among 196 patients analyzed, “Allergy only” was the most prevalent cause of eosinophilia (n=82), followed by infectious diseases (n=41), gastrointestinal conditions (n=26), rheumatologic conditions (n=19), hematologic disorders (n=11), and combination/other causes (n=17). Severity of eosinophilia based on absolute eosinophil count trended higher in younger patients. A statistically significant association was observed between cause of eosinophilia and patient characteristics such as IgE levels, flow cytometric abnormalities, inflammatory markers and CBC abnormalities. IgE levels were elevated in all the study participants (N = 10) in the gastrointestinal group (p=0.04). Flow cytometry abnormalities were notably present in the hematologic and combination/other groups (both 80% abnormal; N = 5, p=0.01). Inflammatory markers were elevated in 68.75% (N = 16) of the participants in the rheumatologic group ( p=0.002). CBC abnormalities outside of eosinophilia were most prevalent in the hematologic group (70%, N = 10, p=0.001). Multinomial logistic regression revealed that each additional year of age increased the odds of eosinophilia due to hematologic (OR: 1.12; 95% CI: 1.01–1.24) and infectious diseases (OR: 1.08; 95% CI: 1.00–1.15). Presence of allergic rhinitis was associated with significantly lower odds of eosinophilia due to infectious diseases (OR: 0.06; 95% CI: 0.01–0.26), rheumatologic conditions (OR: 0.20; 95% CI: 0.05–0.76), and combination/other causes (OR: 0.24; 95% CI: 0.06–0.94) as compared to allergy only being the cause of eosinophilia among these patients. In a separate analysis of 277 patients with complete data, a family history of atopy was associated with higher odds of asthma (OR: 3.12; 95% CI: 1.67–5.89), eczema (OR: 6.33; 95% CI: 3.50–11.77), food allergies (OR: 2.44; 95% CI: 1.12–5.29), and allergic rhinitis (OR: 3.56; 95% CI: 1.94–6.60). Additional CBC abnormalities were associated with lower odds of allergic rhinitis (OR: 0.38; 95% CI: 0.19–0.73).

Conclusions: Allergic conditions are the predominant cause of eosinophilia in pediatric patients. Age and presence of allergic rhinitis are significant predictors of eosinophilia etiology. Elevated total IgE levels, flow cytometry abnormalities, and inflammatory markers are associated with non-allergic causes. A family history of atopy increases the likelihood of atopic conditions, while CBC abnormalities may indicate alternative diagnoses and warrant hematologic evaluation. These findings will support the development of a clinical algorithm designed to expedite the involvement of appropriate consulting services.

This content is only available as a PDF.
2025
Sign in via your Institution