Big IT challenges ahead for precision medicine
While interest grows in the use of genomic data to improve care for patients, the scope of the challenges facing the healthcare industry for personalizing medicine is breathtakingly large.
The promise of genomics in care has been thrust into the spotlight over the past year, thanks to the Obama Administration’s launch of the Precision Medicine Initiative. While $215 million has been budgeted to support these efforts, much work lies ahead before personalized medicine becomes a reality for patients at their local hospitals.
Hospital IT executives believe it will be years before genomic data will affect their electronic health records systems. A recent poll of executives indicates most hospitals and health systems are not yet planning on leveraging such advances in genomics and data analytics to personalize patient care.
The survey, conducted by analytics vendor Health Catalyst, found that 59 percent of respondents indicated that precision medicine will not play a significant role in their organizations over the next five years. Non-academic hospitals and health systems are less likely to expect near-term participation in precision medicine. The survey found that 68 percent of those providers responded that precision medicine will play a "small or non-existent role" in their organization.
Despite the amount of work that lies ahead, thought leaders in the industry are beginning to build the IT foundation that will support precision medicine within healthcare. And some leading organizations already are dabbling in genomics, envisioning how the benefits could be incorporated in patient care.
As work starts at the federal level to launch precision medicine, standards experts already are considering how the information can be brought into patient records, which will be crucial for making genomics part of patient care.
Efforts are underway in several areas to advance the work on standards, says Grant Wood, senior information technology strategist for the Clinical Genetics Institute of Intermountain HealthCare. “There are different drivers going on today, such as consumer empowerment, self-reported data and the precision medicine initiative, as well as genetic care coordination initiatives,” Wood says.
HL7 has developed a data transmission standard that would enable the process of getting genetic information on family health history into electronic health records, as well as standards to transfer genetic variations into records, Wood says.
Under the auspices of the Institutes of Medicine, a collaborative has been formed to investigate challenges of getting genetic history data into electronic health records, with participants including healthcare provider systems, lab companies and vendors. For example, Intermountain is partnering with one of its reference labs and an electronic health records vendor to develop messaging for pharma genomic tests.
Standards organizations need to focus on how to improve care coordination with the use of genomic data, such as determining “the entry points for where that data comes into the healthcare continuum for the patient, and how is it going to be used,” Wood says.
Data sharing will be one of the driving forces that will empower precision medicine efforts, says Howard Levy, M.D., assistant professor of medicine at Johns Hopkins School of Medicine. “It’s sad there’s not a better approach to develop some really rich networks for sharing this information.” There has been a lack of economic incentives to enable the sharing of genetic information, as well as the lack of technical capability to achieve it.
“Medicine is slow to change, but we want to do right by the patient,” Levy says. “As the harm of having data in silos becomes more evident, we believe things will improve in this area.”
Even though the Precision Medicine Initiative is still in its infancy, work has begun on crafting the privacy rules that will govern it. While the initiative’s form and functions are a work in progress, the HHS Office for Civil Rights has crafted some overarching principles to govern how it will oversee the privacy challenges posed by PMI, says Deven McGraw, OCR’s deputy director for health information privacy.
OCR has developed 29 principles “grounded in 10 fundamental assumptions of how PMI will operate,” says McGraw. “A robust data security framework will be built in from the start. This is a new model for scientific research, but it is not widespread.” OCR is developing its first set of “frequently asked questions” regarding privacy and the PMI, and expects to release it soon, she adds.
One of the goals of the PMI is to build a cohort of one million Americans who will agree to contribute their genomic information. “They’re called participants, not patients, and are considered to be partners in the research, with a seat at the table,” McGraw says. The model will be transparent, with the goal that participants will continue to stay involved in the project.
Key to gaining buy-in from one million participants is ensuring that the data will be secure, that they’ll have access to their information and the findings of the research, and that results could be directed to researchers chosen by the participants, McGraw says.
The best way to get public benefits of this is to put patient information back under the control of the patients. There is a way to have privacy and the common good.
New questions arise when genetic data is involved, particularly when it is part of large data sets. “We’re going to solicit input on whether genetic data is protected health information when it’s present in a data set,” McGraw says. OCR further needs to address concerns that participants be fully informed about rights to their data when they sign up to participate.
Public concerns about the extent to which their data will be protected looms large, says Deborah Peel, M.D., founder of Patient Privacy Rights, an organization that advocates for patient control of personal health information. “The public is very concerned about how their health information is being used,” she adds.
Data protection worries could have a chilling effect on citizen participation in the PMI, Peel contends. “You can’t treat human data the same way that you treat other data,” she says. “The best way to get public benefits of this is to put patient information back under the control of the patients. There is a way to have privacy and the common good.”
Even though precision medicine is new to most Americans, several initiatives are under way to accelerate the use of genomic data.
For example, the Food and Drug Administration has launched a cloud-based portal enabling researchers to work together to improve the quality and accuracy of next-generation sequencing (NGS) genomic tests, which compile a vast amount of data on a person’s exact sequence of DNA. The platform, called precisionFDA, is described as a research “sandbox” for the genomics community, where scientists and test developers can explore NGS methodologies to develop standards for the testing technology.
Partners HealthCare is examining the challenges of getting from genetic data to clinical action, realizing that it will require some significant support from information technology. Leaders there say personalized medicine will pose major challenges to electronic health records systems, analytics approaches and systems that provide clinical decision support.
Partners is tackling the challenges through its personalized medicine initiative, in which it is accelerating the use of genetic and genomic information in the clinical setting to improve patient care. Doing that requires a significant transition for information systems that manage data and enable intelligent reporting combined with expert analyses.
Partners has developed GeneInsight Suite, an IT platform to help address challenges of clinical adoption of personalized medicine – aiming to streamline clinical testing, better manage the vast amounts of data generated in genetic testing, support the clinical interpretation process and more easily channel information to clinicians providing care.
“Partners has a long relationship with laboratory vendor Sunquest. These kinds of relationships are always important in establishing extensive partnerships of this type. But there are two other driving factors in addition to the relationship,” says Samuel "Sandy" Aronson, executive director of IT at Partners Personalized Medicine. The first is instances in which GeneInsight can serve different institutions that can be federated with one another to form knowledge-sharing networks. The power of these networks is proportional to the number of nodes sharing data. Because Sunquest has such a large customer base, they have the potential to expand the network far quicker than we could hope to do alone.
“The second is that we see GeneInsight as a continuous learning platform focused in the genetics space,” Aronson adds. “For personalized medicine to reach its potential, it must extend continuous learning processes to include clinical data from other areas. The integration of GeneInsight with Sunquest’s pathology products could accelerate this process in exciting ways.”