Electronic medical records (EMRs) are becoming an increasing important source of phenotypic information for clinical and genomic research. Researchers create and iteratively refine phenotype algorithms using structured and unstructured data to achieve high positive predictive values to identify true cases and controls from EMR data. The Phenotype Knowledgebase website, PheKB, is a collaborative environment to building and validating electronic phenotype algorithms. PheKB was functionally designed to enable such a workflow and has purposefully integrated tools and standards that guide the user in efficiently navigating each of these stages from early stage development to public sharing and reuse. PheKB has tools to enable cross-site collaboration for algorithm development and validation.
Phenotype algorithms can be viewed by data modalities or methods used:
- ICD and CPT codes
- Vital Signs
- Natural Language Processing