WEDI: EHR support for genomic medicine lacking
Genomic medicine requires high-quality data that can be readily accessed and applied in the patient care setting. Yet, despite improvements to electronic health record systems, much work remains to optimize their ability to support genomically informed care.
That’s the contention of a new white paper from the Workgroup for Electronic Data Interchange.
Written by WEDI’s Genomics Workgroup, the document asserts that for the potential of genomic medicine to be realized a computer-based infrastructure must be built to harness the power of clinical data linked with molecular data, to begin storing and utilizing genetic and genomic data from a centrally-managed resource.
“The construction of these ‘Omic data repositories will be key in support of genetic-based care coordination,” argues the white paper. “With the ‘Omic repository playing a central role, the key requirement is to link the data to the electronic health record.”
But, Grant Wood, co-chair of the WEDI Genomics Workgroup and senior health IT strategist at Intermountain Healthcare’s Clinical Genetics Institute, warns that EHRs are not able to effectively capture, share, organize and analyze genomic information.
As an example, Wood notes that in the current computing environment if a physician orders genetic testing for a patient and gets results back from a lab, it is typically returned in a Portable Document Format (PDF). The problem, he says, is that EHR systems do not store structured or coded genetic data.
“We’re working with different vendors and encouraging them to develop this capability,” he adds. “Many times vendors come back and say ‘we’re happy to do this but we’re waiting for our customers to demand it.’ ”
However, the ‘Omic data repository linked to the EHR is merely one component required to build a comprehensive clinical genetic/genomic infrastructure.
According to WEDI, other components include:
- Data pipelines run by bioinformatic tools that capture raw data from the sequencing machine to create clinically useful variant data files (called VCF files), which are specific to the person.
- Hadoop data lakes to store the larger pre-VCF files (BAM/SAM and FASTQ) for future research and analysis.
- Databases that go beyond DNA storing variants that manage single-nucleotide polymorphisms, insertions/deletions, copy-number variants, alleles, and wholegenome/whole-exome sequences, but also other data like protein sequence, gene expression, methylation, and epigenomic data.
- Interface and data integration capabilities that can link to data from multiple sources (sequencing machines, labs, EHRs, clinical data warehouses, etc.).
- Analytic capabilities that can be integrated as modular tools first for annotation, then clinical interpretation and application of genetic information for targeted patient treatment.
- External applications of all kinds to find, display, understand, and act on this data, for doctors, patients and researchers, made possible via application programming interfaces (APIs).
“When a clinician orders a drug, needs genetic information for a diagnosis, or a risk analysis to begin early screening, the EHR can query the repository and get the genotypic data it needs and copies the data locally,” states the white paper. “In fact, with just these two components and one clinical genetic interpretation service, the basics of precision medicine are supported. Standardized data models, standards-based data transmission and APIs are being developed and adopted by many system vendors to achieve this goal.”
Nonetheless, a poll of healthcare executives released earlier this year revealed that most hospitals and health systems are not planning on leveraging advances in genomics and data analytics to personalize patient care. In addition, the poll showed that 63 percent of those surveyed reported that their organizations had no plans to integrate genomic data into their EHRs.
“I’m not aware of a single health system today that has a complete and comprehensive genetic services program in operation,” comments Wood.
Still, 50 percent of the survey respondents indicated that DNA sequencing—the source of genomic data—could have a positive impact on their organizations’ patient treatment strategies.
It’s a sentiment shared by the WEDI Genomics Workgroup which makes the case that genomic data can be leveraged to better coordinate care and achieve better healthcare outcomes at both individual and population levels.
“Accurate and reliable data must be available to those who can use it to improve care and health outcomes,” concludes the white paper. “This includes the consumers themselves (or patients, if seeking care) and their families and caregivers, if appropriate.”
Ultimately, Wood predicts that in the future healthcare will see whole-genome sequencing for all patients, which will become standard care, requiring massive data storage capabilities. The WEDI white paper points out that just one file of genomic code is very large, but a whole genome sequence can take 150 gigabytes or the equivalent of 100 feature length movies.