The United States Immunodeficiency Network (USIDNET) is an NIH-funded research consortium that advances scientific investigation in the field of inborn errors of immunity (IEI). Capturing demographic diversity within the patient registry is important for improving research on IEI and supporting health equity in clinical immunology. Formerly, registry data were collected via an opt-in system with patient informed consent. Now, the registry uses semi-automated de-identified data extraction from EPIC with a consent waiver. This study was performed to assess whether the new ascertainment method improved diversity using data from one site.
De-identified data contributed from CHOP was reviewed from both the old and the new USIDNET registries. Race, ethnicity, sex, and age within the new (n = 1,145) and old (n = 551) registries were defined. Patients without data for a given demographic (i.e. race) were excluded from the analysis for that demographic. Chi-squared and Fisher’s exact test were completed using GraphPad Prism to determine statistical significance.
The racial breakdown of the new registry was: 0.18% American Indian or Alaska Native, 2.55% Asian, 6.94% Black, 72.93% White, and 17.40% other or more than one race. This was a significant (p < 0.0001) difference in racial distribution compared to the old registry: 0.78% Asian or Pacific Islander, 10.85% Black, 86.05% White, and 2.33% other or more than one race. Relative to the old registry, the new registry had a significantly higher proportion of patients from racial minority backgrounds for the diagnosis of agammaglobulinemia (p = 0.0194), but not for severe combined immunodeficiency (p = 0.3937) or chronic granulomatous disease (p = 0.7305). The new registry trended toward a younger median age: 17 years old (0-54) versus 28 years old (9-72).
The design of the new USIDNET registry may better capture a greater representation of patients from racial minority backgrounds compared to the old registry. However, the proportion of patients from racial minority backgrounds within the registry continues to be lower than the populations seen in either total or immunology department CHOP patient visits. While there are many contributing factors, our data may indicate disparities in accessing tertiary care or receiving an IEI diagnosis.
