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

SMART on FHIR application embedded in the electronic health record: HPO term extraction by ClinPhen from clinical notes, followed by clinician curation

Pheno+ is a SMART on FHIR application embedded within the University of Utah Health electronic health record. It retrieves clinician-selected clinical notes and applies ClinPhen, a natural language processing tool that proposes candidate HPO terms, which clinicians then review and curate within the record. In an eight-case crossover design against the standard workflow (physician chart review followed by HPO term selection by a molecular geneticist), Pheno+ reduced mean total workflow time by 66%, from 31.1 to 10.5 minutes (p < 0.001). In a separate cohort of 17 patients with confirmed monogenic diagnoses, Exomiser ranked the causative gene first in all 17 cases using Pheno+-derived HPO terms.

Synthesis written by Geno'X. For the full original abstract, please refer to the source publication.

Preprint, not peer reviewed. Structured phenotyping remains a bottleneck of exome or genome analysis, and embedding it in the record with clinician curation is the right architecture. Sample sizes are very small, and the 17 out of 17 concerns already solved diagnoses: the abstract compares Exomiser with and without phenotype input, not against HPO terms from the standard workflow.

Analysis by Dr Thibaut Benquey