Photo Credit: ALIOUI Mohammed Elamine
A machine learning algorithm can identify patients with common variable immunodeficiency disease (CVID) using their EHRs, according to a study in Science Translational Medicine. Due to the low prevalence and extensive heterogeneity in CVID phenotypes, resulting in delayed diagnoses and treatments, Ruth Johnson, PhD and colleagues presented PheNet to identify patients with CVID. PheNet learns phenotypic patterns from patients with CVID and ranks patients according to their likelihood of having the disease. Researchers found that more than half of patients with CVID could have been diagnosed one or more years earlier with PheNet. In a large EHR database, 74% of the top 100 patients ranked by PheNet were highly probable to have CVID. Researchers validated PheNet with more than 6 million records from medical systems in California and Tennessee.
