Many people view precision medicine as healthcare’s next frontier. The field promises a future where treatments are tailored to each patient’s unique genes, biology and disease risk, with the goal of improving health outcomes and saving lives.
But for health systems, turning that vision into real-life clinical workflows has proven far more difficult than the science itself.
A new report from UPMC’s Center for Connected Medicine showed that the field is making progress, with more provider organizations than ever before building out formal precision medicine programs. In fact, it found that more than three-quarters of U.S. health systems now operate these programs for genetic testing and personalized care. However, the report also highlighted the hurdles many organizations are still struggling with as they try to scale these programs, including reimbursement, data integration and patient engagement.
Experts agree these obstacles need to be addressed before precision medicine can graduate from a niche offering to a standard component of care delivery.
From promise to implementation
UPMC’s report highlighted that the precision medicine field has come a long way in the past decade — particularly in areas like oncology, pharmacogenomics and maternal health, in which hospitals are increasingly integrating genetic data into clinical care.
Many health systems don’t consider their precision medicine programs new anymore, with 35% of health systems reporting their programs are at least five years old. The report also found that about 60% of health systems have a designated individual overseeing these efforts, such as a director of precision medicine or chief genomics officer.
But when looking at the report’s more qualitative data — interviews with health system leaders — it is clear that these health systems are facing roadblocks when it comes to scaling precision medicine programs across their organization. The main obstacles are reimbursement challenges, physician education and integrating massive genomics datasets into existing EHR systems, said Adrian Lee, director of the Institute for Precision Medicine at UPMC.
In the past, genomics programs typically operated in isolation because the healthcare system couldn’t handle large, complex genomic datasets. As EHRs and lab integrations have improved, more hospitals are working to transform genomic results from PDFs, faxes and standalone lab portals to structured data elements — which means providers are starting to be able to embed genomics directly into EHR workflows, Lee explained.
He noted that there’s still a lot of work left to be done — and that the more providers work on these integrations, the more they can enable AI-driven insights and scalability.
There’s only so much scale you can reach without adequate reimbursement, though.
Reimbursement for many precision medicine services remains inconsistent, with coverage varying widely by payer, indication and test type.
Some targeted therapies and genetic tests are reimbursed, usually in oncology. However, many broader applications — like population genomic screening and much of pharmacogenomic testing outside oncology — face limited payment pathways, often requiring prior authorization or strong supporting evidence of clinical utility, Lee explained.
In oncology, several genomic tests — including Foundation Medicine’s FoundationOne CDx and Natera’s Signatera minimal residual disease test — are widely reimbursed, particularly when tied to cancer treatment decisions.
Payers are more likely to push back on broader use cases like population-level genomic screening or pharmacogenomic testing outside oncology because the evidence of clinical utility and cost savings is still developing, Lee said.
He said there is a growing push to generate stronger evidence for precision medicine’s clinical and financial value.
Payers want clearer proof that these approaches improve outcomes and reduce costs before they decide to broadly reimburse them, which means health systems have to collect better data. They need to produce data showing benefits like earlier diagnoses and free hospital admissions, Lee said.
He pointed to whole genome sequencing in NICUs as a compelling example for which he thinks providers should be collecting more data.
When an infant enters the NICU, there is usually a long, expensive diagnostic odyssey. But if infants were to receive genome sequencing upon their arrival, the care team may be able to quickly identify their underlying conditions — leading to faster diagnoses, more targeted treatment and potentially lower costs, Lee noted.
“About 40% of the time, [genome sequencing] will identify, particularly in those who have an underlying genetic disorder, exactly what’s wrong with them. And once you have the diagnosis, you have potential therapeutic options,” he remarked.
Making genetics usable at the point of care
For Vanderbilt University Medical Center in Nashville, one of the biggest challenges in scaling precision medicine has been figuring out how to operationalize genetic testing inside its everyday clinical workflows.
The health system has made huge strides after spending years working to integrate genetic test ordering and results directly into its EHR.
Less than five years ago, every genetic test had to be ordered via a paper form, said Alex Bick, director of genetic medicine and clinical pharmacology at Vanderbilt.
“You would have to fax that paper form to the lab. You’d have to then enter what you put on the paper form into a website that the lab testing company ran, and then you had to send in a special tube. This whole process was really challenging, and made it so that really the only people who could do genetic testing were genetics clinics and genetic counselors who were highly trained to jump through all these hoops,” Bick explained.
To solve this problem, Vanderbilt worked with Epic to develop a genomics module that allows clinicians to order genetic tests directly through the EHR, just like they would with any other lab test. The system also returns results as structured data integrated into the patient record, which Bick said makes genetic testing a lot more manageable for clinicians.
“That has made a huge difference for scaling because now any clinician at Vanderbilt who would like to order genetic tests can do it, and there’s no friction in the actual act of getting the testing done and getting the results back,” he declared.
Building the right infrastructure within the EHR has also helped Vanderbilt use genomic data for population health management and preventive screening, Bick added.
As genomic data becomes part of the clinical record, health systems are beginning to use AI to spot patterns that might signal undiagnosed hereditary diseases, he said.
At Vanderbilt, researchers are developing algorithms that identify patients who may have these undiagnosed conditions and should receive testing. Bick also pointed out that health systems are exploring how to use AI to help explain complex genetic test results to patients.
He noted that patients are increasingly arriving with genetic information from direct-to-consumer testing services and whole genome sequencing companies — which is creating additional pressure on providers to develop the infrastructure and expertise needed to interpret and act on these results.
Moving genomics to the prescription pad
At Endeavor Health in the Chicago area, precision medicine leaders are tackling a more operational issue: how to move from genetic insight to real-time prescribing decisions.
Peter Hulick, chair of precision medicine, said the health system’s early focus has been building EHR infrastructure that can turn pharmacogenomic data into point-of-care guidance for clinicians.
That includes integrating pharmacogenomic results directly into the EHR so they are visible at the moment a clinician is prescribing, rather than buried in separate reports or other systems. In some cases, these tools are built into order sets that automatically suppress or remove medications when a patient’s genetic profile suggests there’s decreased efficacy or a higher risk of adverse effects, Hulick noted.
The goal is to enable greater scale by making it easier for clinicians to adopt genomic workflows. Many clinicians still lack confidence using genetics in routine care because this testing has historically been handled through specialized gatekeepers like genetic counselors, Hulick stated.
As of this spring, Endeavor has embedded genomic data for more than 50,000 patient medical records. As the health system continues to scale this program, Hulick said it will rely on a “learning health system” approach.
“I like to call it a ‘genomics learning health system’ — where we implement, we look to see if we are achieving our goal, we reflect on it and try to decide if this is how we intend to work. And then, we learn from it to make that cycle better. And we keep going, and then you naturally start to get that synergy,” he declared.
In his eyes, early metrics like test volume aren’t sufficient on their own. Meaningful outcomes such as earlier cancer detection and reduced disease burden take years to fully materialize.
While early signals like improved screening uptake and stronger patient engagement are emerging, Hulick said he is excited to see more rigorous, long-term analysis start to take shape over the next few years.
“One of the challenges for a health system is to keep people engaged in their health. We are seeing those signals, and we’re doing the work to try to make our data analysis more robust so it can reach the level of publication-quality. But we’re not there yet,” he remarked.
Specialty service or standard care?
If I remember correctly Geisinger Health System was doing universal gene testing for free. They announced this several years ago. Can we provide an update on this program? I saw you talked to them below, which is great.
Geisinger in Pennsylvania was one of the earliest adopters of large-scale precision medicine.
This began with its MyCode Community Health Initiative in 2007, which was one of the first large-scale programs to return genomic sequencing results to patients, as well as integrate them into routine clinical care, noted Christa Lese Martin, Geisinger’s chief scientific officer. The health system later added genomic sequencing through a collaboration with Regeneron Pharmaceuticals in 2013.
The program now includes more than 370,000 consented participants and genomic sequencing data on roughly 230,000 patients — which Martin said makes it one of the world’s largest health system-linked biobanks.
Geisinger is currently working to transition genetics from a specialized field into routine healthcare delivery. Martin said precision medicine can no longer remain confined to genetics clinics because genomic information is increasingly relevant across oncology, cardiology, neurology and many other specialties.
As researchers uncover more links between genetic variants and disease, Geisinger is focused on helping specialists use genomic information more seamlessly to guide screening, diagnosis and treatment decisions, she explained. In many cases, genetic testing can now determine whether a patient qualifies for a targeted therapy or faces elevated risk for a serious condition.
Martin pointed to newborn genomic screening as an example of where the field might be headed.
Instead of ordering separate genetic tests throughout a patient’s life, some researchers are exploring whether every baby could receive genome sequencing at birth. The idea is that the information would be stored and reused over time to inform preventive care, diagnose diseases earlier and make the patient’s care plan more personal.
At Geisinger, the genomics program has been steadily growing for nearly two decades, Martin noted. About 230,000 patients have now received genomic sequencing through the health system — but she pointed out that scaling these programs across an enterprise remains difficult, with many organizations still working through operational and reimbursement challenges.
Professional guidelines play a major role in reimbursement adoption — and payers are more likely to cover genomic testing once major medical societies establish clear clinical guidelines showing the tests improve care, Martin stated.
Like UPMC’s report, she said providers need to continue diligently working to collect and publish this evidence so that precision medicine can move beyond isolated pilot programs.
“There’s a lot of movement in those areas, but there still needs to be a lot of coordination,” Martin remarked.
Fixing the final mile
Patient drop off is another pain point that precision medicine professionals are figuring out how to best address.
Patients can disengage or fail to follow through at multiple points in the care pathway, pointed out Daryl Pritchard, science policy consultant at the Personalized Medicine Coalition, a nonprofit seeking to advance the field.
For instance, patients can miss follow-up testing appointments or fail to retrieve the therapies tailored to their genetic profile, he explained.
In lung cancer, many patients still never receive the biomarker-driven treatments that are best matched to their disease. Pritchard said about only about 35% of lung cancer patients end up actually receiving their personalized treatment.
“That’s pretty remarkable in lung cancer, because there’s a very robust personalized medicine pathway in lung cancer,” he added.
Pritchard said improving coordination, testing workflows and clinical decision support will be key to eliminating these gaps. He noted that these challenges are especially pronounced in community hospital settings, where resources and infrastructure are more limited than at large academic medical centers.
He and the other experts interviewed for this article agree that the next phase of precision medicine depends less on new scientific breakthroughs and more on helping health systems identify scalable implementation strategies that can work across different care environments.
Photo: Josh Hawley, Getty Images
