By Matt Yuill, M.D., Vice President, Quality and Risk Analytics, Interpreta
Genetic testing has already been influencing clinical decision-making. A number of single gene tests are available to evaluate breast cancer risk or determine how long a patient with deep vein thrombosis will need to stay on a blood thinner. Soon, clinicians will be incorporating information from significant portions, if not all, of a patient's entire genome into many aspects of clinical decision-making. A recent symposium on precision medicine at Mayo Clinic (Nov. 2017) suggested a clinical goal of having genomic information “preemptively” available in the EHR so that it could be seamlessly incorporated into clinical workflow without having to wait for genomic results.
Genomic information will inform medication safety and treatment efficacy at the point of prescribing, help predict risk of disease and disease progression, and help predict individual response to a particular treatment or surgery. Genetic information will not be just another lab result, rather it will intercalate through all parts of the patient's electronic health record, as part of the prescribing workflow, problem list, medication list, risk calculators, and can be consulted as needed to weigh treatment options, and even for population health functions.
The ever-decreasing cost of next-generation sequencing (NGS) and increasing consumer demand for genetic information make it likely that clinicians will soon be obtaining a patient's whole genome at once rather than ordering and interpreting one genetic test at a time. The cost for whole genome sequencing has recently fallen under the $1,000 mark and companies like Illumina are looking to get the price down to $100 per genome.
With costs soon to be removed as a barrier, other challenges will be storing genomic information and analyzing it as needed. Depending on how much metadata is stored, it can take anywhere from two to 60 Gigabytes to store just one patient's genome. EHRs will not be able to store it, and networks would not be able to handle downloading it from cloud storage on demand. EHRs and clinical decision support (CDS) systems will, therefore, have to be smart about pulling down or storing just the relevant variant information when needed for CDS.
Analytics will need to comb through thousands of known variants associated with disease along with hundreds of lab, medical, pharmacy, and other data points on a single patient and serve up relevant information at the point of care. The clinical relevance of genetic variants will be constantly evolving as new research is published and new guidance is released by the FDA and leading genetics organizations, necessitating frequent reanalysis, especially in light of changes in a patient's medical history.
Health plans, policy makers, and providers will be learning over time where it will make sense to leverage genomic information. Pharmacogenomics will likely be one of the leading use cases for whole genome or exome (coding regions) sequence information. So far, the FDA has recommended genetic testing for over 100 medications to determine either drug safety or drug effectiveness in the presence of specific genetic variants. One example is Warfarin, a drug that some patients do not break down as well, potentially causing levels to build up more quickly increasing the risk for a bleeding event. In addition to drug safety, having genetic information already available at the point of prescribing can help providers avoid prescribing new medications that are less effective in some people. This could reduce time wasted on less effective therapy and get patients on the right medication, and better, faster.
Beyond the better-defined use case of pharmacogenomics, it may take a little more time to incorporate a patient's genomic profile into other aspects of clinical decision-making. Genomics may improve risk prediction for a disease or treatment outcome much better than the standard clinical risk factors such as family history. Genomic profiles will be another aspect of a patient's chart that clinicians will evaluate beyond just medical history, medications, and lab results. Genomic profiles will be consulted for new medication decisions, treatment decisions, and personalized cancer screening decisions. But, in addition to interpreting more information, it will be incumbent on clinicians to generate more information as well. More precise and comprehensive documentation of patient attributes and clinical histories will be required to better inform genome-wide association studies (GWAS) to uncover new genetic associations.
So, the integration of genomics into EHRs, clinical workflows, and clinical decision-making will be no small task and will be constantly evolving over many years to come. Clinicians and IT departments will need to be adaptive as these changes occur, as we all collectively figure out how to leverage genomic information to improve individual and population health.
About The Author
Matt Yuill, M.D. is a physician and Vice President, Quality and Risk Analytics at Interpreta. His company is a provider of a real-time analytics engine that continuously updates, interprets, and synchronizes clinical and genomics data, creating a personalized roadmap and enabling the orchestration of timely care. It is online at www.interpreta.com. For correspondence, please email firstname.lastname@example.org.