Projects
Current projects in the Informatics of Genomic Medicine Laboratory use pharmacogenomics as a starting point to develop the tools and technologies needed to realize genomic-based precision medicine.
Developing next-generation genomic clinical decision support
Traditionally, genomic-based clinical decision support rules have been implemented as pop-up alerts that interrupt the clinical workflow. In addition, the clinical decision support rules usually consider only a single gene at a time.
As clinical-grade whole-exome and whole-genome sequencing becomes more common, technologies are needed to interpret patient genomic data and deliver comprehensive, integrated knowledge to clinicians at the right time. In partnership with the Center for Individualized Medicine and the Office of Information and Knowledge Management, we are developing novel systems that provide scalable next-generation genomic clinical decision support. In collaboration with the Medication Safety Code initiative, we are exploring technologies that will enable patients to access personalized pharmacogenomics recommendations using mobile devices.
Standardizing clinical genetic data
Genomic data can't be used to improve health if it can't be accurately communicated and correctly understood by both humans and computer systems. Standards for data representation and exchange are necessary to bridge this gap and enable genomic clinical decision support. However, existing standards are incomplete and not universally adopted.
These limitations form a barrier to the efficient exchange and use of genomic information by clinicians, researchers, and electronic systems, including clinical decision support services. This area of research aims to address these limitations by improving existing standards for the representation of clinical genetic data.
Collaborative partners include:
- Pharmacogenomics Research Network
- Clinical Pharmacogenetics Implementation Consortium (CPIC)
- HL7 International
- Clinical Genome Resource
Enhancing standards for medications
Standardized representation of drug information is needed to facilitate the management of medication lists for clinical decision support rules, including those used for clinical pharmacogenomics. The existing standards used to represent drug data within clinical information systems do not contain the level of precision necessary to efficiently maintain large-scale genomic clinical decision support rules. This study, in collaboration with Mayo Clinic Pharmacy Informatics, systematically identifies and resolves limitations in drug terminologies.
Capturing and disseminating pharmacogenomic clinical decision support knowledge
While all medical centers use the same body of knowledge to inform the design of pharmacogenomic clinical decision support, no two implementations are the same. Sharing detailed information about pharmacogenomic clinical decision support implementations is critical to establishing standards of care and enabling multisite outcomes studies. Yet no conventions exist. In fact, one of the greatest challenges in comparing pharmacogenomic clinical decision support rules is unambiguously documenting how a genetic test result is generated, interpreted, and acted upon clinically.
In partnership with the Pharmacogenomics Research Network Translational Pharmacogenetics Project (PGRN-TPP), the Clinical Pharmacogenetics Implementation Consortium (CPIC), and the Electronic Medical Records and Genomics (eMERGE) Network, we are developing novel methods for capturing pharmacogenomic clinical decision support knowledge and creating public repositories for sharing.
Developing portable pharmacogenomic clinical decision support algorithms
Clinical pharmacogenomic guidelines are often used as the basis for genomic clinical decision support rules, the development of which is very labor-intensive. Currently, each adopting site must interpret the guideline content individually because electronic medical record systems use proprietary syntaxes for clinical decision support rules. This can result in variations in clinical decision support logic among sites, leading to inconsistencies in patient care and complicating future outcomes studies of pharmacogenomic-guided therapy.
Given the pace at which existing guidelines are updated and the rate of publication of new pharmacogenomic guidelines, the current process of disseminating pharmacogenomic knowledge and translating it into clinical practice is not scalable. We are developing methods that use emerging standards to enable the generation and dissemination of pharmacogenomic clinical decision support algorithms that can be shared across electronic medical record systems. This work is done in collaboration with Arizona State University and HL7 International.
Evaluating the outcomes of genomic clinical decision support
It is necessary to scientifically evaluate genomic clinical decision support implementations to quantify their impact on patient care. In partnership with the Mayo Clinic Clinical Decision Support Team and the Translational Pharmacogenetics Project within the Pharmacogenomics Research Network, we are extracting and curating electronic medical record data related to pharmacogenomic outcomes. This project will provide critical feedback on our existing pharmacogenomic clinical decision support implementations, and it will inform the development of future genomic clinical decision support interventions.