Implementation best practices: The best approaches to precision medicine
Precision medicine is just starting to gain ground in the healthcare industry. Most patients have yet to have medical care designed to enhance effectiveness or therapeutic value for select groups of patients, especially by utilizing genetic or molecular profiling.
But some forward-looking healthcare provider organizations have started using precision medicine in their day-to-day operations. And the number of technology vendors with precision medicine offerings is growing.
Precision medicine is indeed the future of medicine overall. And as a result, it is time for all healthcare provider organizations to start thinking about how they can incorporate the high-tech offering.
Here, executives from precision medicine groundbreaking provider organization NorthShore University HealthSystem and precision medicine technology vendors 2bPrecise, Orion Health and Translational Software offer healthcare CIOs and other leaders and IT workers key best practices and tips for implementing precision medicine technologies.
Future-proofing technology and strategy
Right from the start, healthcare provider organizations must future-proof their precision medicine strategy, said Assaf Halevy, founder and CEO of 2bPrecise.
“One of the riskiest aspects of the CIO’s job is investing in advanced technology to support organizational needs – needs that evolve and change over time,” he said. “This risk is particularly acute when selecting an informatics strategy to support a highly dynamic aspect of clinical care, such as precision medicine.”
Healthcare leaders are grappling with the challenge of how they can leverage insights from genomic science to improve the care they deliver to patients – now and into the future.
“CIOs must define a sound genomic data management strategy and activate an execution plan that engages users across the entire enterprise,” Halevy stated. “CIOs can help drive success by making sure they select a precision medicine platform that meets four specific criteria.”
"One of the riskiest aspects of the CIO’s job is investing in advanced technology to support organizational needs – needs that evolve and change over time."
Assaf Halevy, 2bPrecise
First, the platform needs to integrate with multiple EHRs, he said. This is critical because most health systems today leverage more than one clinical information system – different software in the ambulatory versus acute setting, for instance, or specialty software for specific clinical disciplines, he explained.
“All areas can use genetic and genomic information, so it’s critical that it be made universally available,” he said. “Likewise, if the organization at some time replaces an EHR system, it wants to ensure the underlying precision medicine foundation continues to support enterprise needs.”
Data from a wide variety of labs
Second, a precision medicine platform needs to consume genetic and genomic test results from a wide variety of molecular labs, Halevy advised.
“As an organization begins its precision medicine journey, it often finds some departments are already doing molecular testing – and each might use a different lab,” he said. “Additionally, some labs offer specialized testing services addressing specific clinical needs – such as oncology or pharmacogenomics. To ensure value is extracted across the entire organization, the precision medicine solution must be able to ingest from a broad range of data sources.”
Third, a platform needs to provide in-workflow access to multiple knowledge databases, Halevy said. Although an individual’s genomic profile does not change over time, genomic science does – and at a dizzying pace.
“Clinicians must be comfortable that their precision medicine solution incorporates the latest research as they factor genomic information into their clinical decision making,” he said.
And fourth, a precision medicine platform needs to deliver flexibility to match clinical priorities, he said. Typically, an organization determines a “starting point” for its precision medicine strategy, such as oncology or behavioral health, and then scales and expands, he said. The precision medicine infrastructure must be agile in supporting these unique priorities, he stated.
Another future-proofing advocate
Dr. Chris Hobson, chief medical officer at Orion Health, agrees with Halevy, saying his first best practice is future-proofing.
“CIOs need to think very carefully about where they are headed with their current project portfolio and what they want precision medicine to do in the next three to five years for their organization,” he said. “Although in general, ‘more data from more sources is better,’ there are some specifics to consider. The field of precision medicine is already rapidly expanding beyond genomics to include evolving and entirely new concepts such as microbiomics, proteomics, metabolomics, phenotypes and endotypes.”
Machine learning is much more likely to be useful and the outputs to be valuable if the input data is sourced from a diverse population, with data elements drawn from across all the critical indicators, he explained. Ideally, these include clinical and the various “*omics” and non-traditional sources such as the social determinants of health, patient generated-data and behavioral data, he added.
“The other part of future-proofing is scalability,” Hobson said. ”The widely varied data drawn from a broad base of patients also represents potentially massive amounts of data that needs to be captured for each individual in a population. To ensure a high degree of accuracy and reliability, CIOs need to look for proven, enterprise-class systems that have also demonstrated the ability to handle massive data sets.”
From a technology perspective, therefore, CIOs should consider looking for technology architectures with the greatest flexibility to handle predictably large increases in both data volume and data types over the next three to five years, he added.
Work closely with strategic decision makers
Another precision medicine technology implementation best practice from Hobson: Work closely with both the provider organization’s domain experts and strategic decision makers.
“In order for precision medicine to work well, a high degree of fundamental knowledge, expertise and experience is needed to come up with breakthrough insights into an organization’s data,” he explained. “This is not something that can be done on a part-time basis by just anyone and potential pitfalls can lie in unexpected places. Technology professionals are very used to the ‘gotchas’ that crop up in any technology implementation. In the case of precision medicine, those unexpected highly complex issues can arise at any point.”
"Many generously funded and well-executed programs have resulted in disappointing utilization because their communities were not ready to adapt clinical practices in order to adopt precision medicine."
Don Rule, Translational Software
Having established the point around a high degree of domain expertise, Hobson strongly recommends as a best practice that CIOs take careful account of the organization’s strategic goals.
“This is important to be sure that the precision medicine work is clearly focused to align with the organization’s overarching strategic direction,” he said. “There are virtually limitless interesting research questions to be tackled with these exciting new technologies. Accordingly, a lack of strategic focus to the work runs the risk that it may result in interesting results that do not advance the organization’s needs as a whole.”
Playing nice with workflow
Halevy of 2bPrecise advised that genomic test results must be made “workflow-friendly.”
“The results from genetic and genomic tests can have a tremendous impact on clinical decision making,” he cautioned. “They can help providers arrive at a highly accurate diagnosis sooner, choose the right medication as the first line of therapy instead of using a trial-and-error approach, and identify targeted therapies for heritable conditions in specialties such as cardiology, neurology and neonatology.”
But to deliver on this value, healthcare organizations must bring genetic and genomic test results directly into the clinical workflow so they are readily available and actionable, he stated.
“Currently, these test results are often returned as paper or scanned documents and saved in the HER,” he explained. “Unfortunately, to be able to use them, clinicians must know that the test was done and results are available somewhere in the clinical information system, leave their normal workflow and hunt for the molecular lab report, figure out what the results mean in context of the patient’s current condition, and interpret and apply the information as they consider other factors in the broader patient record.”
This time-intensive, manual approach is a barrier to clinicians’ ability to leverage valuable genomic data, Halevy said.
“Forward-thinking CIOs are in a position to accelerate adoption of precision medicine – and greatly improve the quality of care their organization delivers – by seeking precision medicine solutions that ingest genomic results as discrete data, integrate them with clinical information, and deliver this combined record to the clinician within the patient context and at the point of care,” he said.
Grounding precision medicine in primary care
NorthShore University HealthSystem, is an early adopter and national leader in personalized medicine. NorthShore has integrated genomic and personalized medicine across its health system, including at the primary care setting. The advanced tools and technologies available across NorthShore present a clear opportunity to strengthen the patient relationship, aid the clinician decision process, improve care and provide value in a variety of ways, said Dr. Peter Hulick, medical director of the Mark R. Neaman Center for Personalized Medicine at NorthShore.
“NorthShore recently made a significant new step forward in integrating and advancing personalized medicine at the population scale via the DNA10K program developed in collaboration with Color,” a precision medicine technology vendor, Hulick said. “DNA10K is a complimentary whole genome sequencing pilot program available for 10,000 eligible patients and is also one of the largest known U.S.-based genomics efforts grounded in primary care.”
"The insight of pharmacogenomics and the ability to integrate into clinical care via our EHR platform has proven to be a key component of our integrated approach and is generating significant value in patient care."
Dr. Peter Hulick, NorthShore University HealthSystem
By integrating the advanced genomic testing platform into the primary care setting, both patients and physicians have access to information for developing potential treatment and prevention plans as part of the ongoing care journey, he explained.
“The Color platform is designed to specifically explore genetic information on multiple cancers – including breast, ovarian, uterine, colon, melanoma, pancreatic, gastric and prostate cancers – genetic forms of heart disease and pharmacogenomics to inform metabolization of medications,” he noted. “Patients participating in DNA10K also have access to Color’s platform of additional genetic insights, such as ancestry, lactose intolerance and caffeine metabolism.”
In delivering results to patients, whether positive or negative, there is a clear benefit in determining risk and mapping a treatment plan in collaboration with their NorthShore primary care physician and advanced care team, Hulick stated.
“Results may also benefit a patient’s family by helping inform and guide risk and awareness of health, as well as the potential need for additional testing, monitoring or action that may be necessary,” he said. “This is another powerful tool to aid NorthShore physicians and support their clinical practice with insights and information for informed decision making.”
Integration with the EHR
Dr. Hulick offered another best practice when implementing precision medicine technologies: Through NorthShore’s advanced primary care model, the organization has bridged the gap between electronic health records and genomic medicine by incorporating patients’ genetic and family history into standard diagnostic assessments.
“The insight of pharmacogenomics and the ability to integrate into clinical care via our EHR platform has proven to be a key component of our integrated approach and is generating significant value in patient care,” he said.
Pharmacogenomics uses genetic testing and analysis to understand how a person responds to drugs – whether they metabolize the drug faster or slower than others based on genetic variants. The implementation of pharmacogenomics on a broader scale assists both physicians and patients, ensuring the right dose of the right medicine, he explained.
“NorthShore launched its Pharmacogenomics Clinic in March 2015, which is among the first of its kind,” he stated. “As a result, the clinic received strong interest from patients and referring physicians. Most referrals have been in the pain management and mental health space. On average, 97% of patients receive actionable information from a pharmacogenomics gene panel (15-20 genes), and more than half have had variants in multiple pharmacogenomic related genes.”
The most common gene variant found depends on the patient’s ethnic background. With the clinic, NorthShore is able to preemptively screen patients for reactions to medications based on their genetic makeup. Clinical staff at NorthShore run tests showing patients their response to a number of drugs based on their genomic makeup.
“This insight is designed to help our physicians personalize care for each patient to support improved outcomes, prevention and overall health,” Hulick said. “By integrating this information into NorthShore’s electronic health records system, analyzing risk factors for hereditary cancers and implementing pharmacogenomics into everyday clinical care, NorthShore has leveraged health technology to provide value at every physician and patient touchpoint.”
Technology is important, culture is crucial
Don Rule, CEO of Translational Software, said that while precision medicine technology obviously is important to any precision medicine program, culture is critical.
“Many generously funded and well-executed programs have resulted in disappointing utilization because their communities were not ready to adapt clinical practices in order to adopt precision medicine,” he said. “Be sure that a critical part of the program is to identify and educate a base of end-users that has both a defined benefit and a financial model for making the program successful. Plan for incremental steps that show real value before investing in ‘moonshot’ technology.”
Another best practice from Rule: Recognize the difference between research and clinical requirements.
“Researchers focus on analyzing patterns across the largest possible cohorts of individuals to advance their field of knowledge,” he explained. “Clinical applications apply pre-existing clinically validated knowledge to individual records to understand the implications for that patient. Where research systems require flexible queries and machine learning to provide novel analyses, clinical systems must supply repeatable, reliable transactions that are validated for clinical use.”
Trying to use a system for an unintended purpose or feeling that one needs to solve both problems with a single system can be overly expensive and under performant, he added.
The minimum viable investments
On another front, healthcare provider organizations must focus the technology plan on the minimum viable investments in new technology to satisfy the clinical goals, Rule advised.
“There is a tremendous amount of new technology entering the market, including Next Generation Sequencing, FHIR, data mining and many other peripheral technologies such as wireless,” he noted. “The fact is that of the billions of bases that are available in a full genome, a remarkably small number of variations have been validated as clinically actionable.”
For this small set, HL7 and relational technology can enable an organization to use the technology and expertise that it has in-house to quickly build a consensus for a program and develop a justification for future technologies – for example, deep EHR integration – that will truly add value, he concluded.