Precision Medicine promises to lower healthcare and improve outcomes. Precision medicine is enabled by the capture of a patient’s genomic data, integrated with their electronic health records (EHRs) for the purpose of informing targeted treatment. However, end-to-end clinical workflows are complex, often necessitating support for custom use cases and organization-specific requirements. Augmenting existing EHR systems with genomic data is challenging, requiring some level of integration with special-purpose cyberinfrastructure. Furthermore, genome interpretation (variant prioritization and reporting of incidental findings) is not a solved problem and platforms must be adjusted to fit maturing methods and recommendations.
To meet these needs for studies being performed at UNC Hospitals, RENCI helped to develop the Clinical Decision Support System (CDSS). The CDSS is a comprehensive framework for automated clinical reporting workflows with integrated genomic analysis using high performance computing (HPC). The framework includes end-to-end support for essential clinical workflow features, including patient intake, sample accessioning, assay monitoring, genomic data analysis, auditing, secure access, clinical reporting and follow-up .
CDSS has been used to support genomic variant interpretation by providing clinicians with a dynamic visual snapshot of the genomic data most relevant to an individual patient. This platform has also been used to capture, store, and curate population-level data in support of biomarker development. To capture patient-level data, such as a list of potentially pathogenic and actionable genomic variants that could inform a patient’s care, the CDSS draws on evidence from published literature and annotation databases. The system also uses a patient’s health status to guide the construction of a variant list. To archive data that could be useful in biomarker development programs, CDSS supports the curation of whole exome sequencing data with extensive metadata (e.g., information on how the data was collected, quality metrics, and provenance).
The genomic variant interpretation feature in CDSS allows integration with medical records, enabling genotype/phenotype analysis and variant prioritization. Variants found via genomic testing are prioritized using the automated update and versioning annotation system in CDSS. The platform also provides the ability to create user views for manual confirmation of variants, for example, after Sanger sequencing confirmation. Confirmed variants are integrated into a formatted patient report that can be exported to a PDF. CDSS also has a feature of end-to-end workflow auditing that allows users and computers to log their action at each step to support security audit and regulatory compliance.
Researchers at RENCI and UNC-Chapel Hill have worked together to develop CDSS, a framework that has been used to build multiple systems to support the entire clinical workflow for analyzing thousands of patient samples. These workflows span from patient enrollment through to report, including support for patient scheduling, clinical notes, genomic variant annotation, analysis monitoring, and system auditing. RENCI’s researchers have also worked on integrating CDSS with high performance computing systems for genomic analysis of sequenced human variation.
CDSS Technical Team
Bradford Powell, UNC School of Medicine
NC GENES project
NC NEXUS project
National Institute on Drug Abuse
UNC Department of Genetics
UNC Information Technology Services Research Computing
The North Carolina Translational and Clinical Sciences Institute (NCTraCS)
Lineberger Comprehensive Cancer Center