The United States Immunodeficiency Network (USIDNET) collects de-identified patient data from hospitals across the country to study inborn errors of immunity (IEI). The registry contains records on >6,000 patients with IEI currently. The longitudinal data have been extracted back to 2018. This registry represents a critical resource for research on IEI. This study was done to evaluate the registry for its data quality.
PHIdentifier runs on a secure, high-performance computing (HPC) environment to efficiently process large volumes of text data to perform a multilayered de-identification process. The model’s responses are combined with rule-based checks to ensure that only sensitive information is replaced with placeholders, preserving all other clinical content. This has allowed a waiver of consent, which has facilitated enrollment. The current registry, as of January 2026, had field counts extracted for this study on data quality.
The registry contains 6,272 patients. 45% are female. The predominant ethnicity was non-Latino, 4,143 (66%). The most frequently enrolled races reported were white (71%), other (8%), and black or African American (7%). Demographic data were found for 100% of enrollees. Social history was found for 88% of subjects, diagnosis for 77%, medications for 73%, immunizations for 69%, and allergies for 56%. The registry contains >1 million medications on 4,485 patients, >8,000 imaging reports on 2,793 patients, and >50,000 laboratory studies on 4,262 patients. These data represent a comprehensive landscape of 6,272 patients with IEI.
USIDNET is a very large registry of patients with robust longitudinal data of varying types, making it a valued resource for the community. Data requests are accepted on a rolling basis, and assistance is available for statistical support for those submitting queries producing large or complex data sets.
