Pheno+: An Electronic Health Record-Integrated, Clinician-in-the-Loop Application for Efficient Phenotyping in Clinical Genomics

Objective To develop and evaluate Pheno+, an electronic health record (EHR)-integrated application that streamlines the generation of structured phenotypes for genomic testing and analysis. Materials and Methods We developed Pheno+ as a Substitutable Medical Applications and Reusable Technologies on Fast Healthcare Interoperability Resources (SMART on FHIR) application embedded within the University of Utah Health EHR. Pheno+ retrieves clinician-selected clinical notes and applies ClinPhen, a natural language processing tool that identifies candidate Human Phenotype Ontology (HPO) terms from clinical text. Clinicians review and curate the resulting HPO terms within the EHR. We used an eight case crossover design to compare total workflow time and HPO term counts between Pheno+ and a standard workflow involving physician chart review followed by HPO term selection by a molecular geneticist. In a separate cohort of 17 patients with confirmed monogenic diagnoses, we evaluated the utility of Pheno+-derived HPO terms for diagnostic gene prioritization by comparing Exomiser performance with and without phenotype input. Results Pheno+ reduced mean total workflow time by 66%, from 31.1 to 10.5 minutes (p < 0.001). Exomiser ranked the causative gene first in all 17 cases (100%) using Pheno+-derived HPO terms. Discussion and Conclusion Pheno+ reduced the time and personnel required to generate clinician-curated HPO term lists while producing phenotypes that provided meaningful diagnostic information for downstream genomic analyses. These findings support EHR-integrated, clinician-in-the-loop phenotyping for genomic workflows.