In Part Two of our round table discussion with members of Project Santa Fe, the conversation turns to the practical aspects of implementing the Clinical Lab 2.0 principles.
From mindset shifts and training requirements, to funding and impact measurement, the panel lays out the key considerations for laboratories in moving towards value-based healthcare.
The panel is:
Khosrow Shotorbani, President and CEO, Project Santa Fe Foundation; Founder, CEO, Lab 2.0 Strategic Services, LLC
James Crawford, Professor and Chair Emeritus, Department of Pathology and Laboratory Medicine, Donald and Barbara Zucker School of Medicine at Hofstra/Northwell, Northwell Health, New Hyde Park, New York; Chair, Board of Directors, Project Santa Fe Foundation.
Ulysses Balis, Chief Medical Officer of the Diagnostic Medicine Consortium and Chief of Computational Pathology at the University of Michigan, Ann Arbor, Michigan
Read Part One of the discussion, about the impact of the Clinical Lab 2.0 movement, here.
How can pathologists translate the strategic vision of Clinical Lab 2.0 into practical action?
James Crawford: Pathology and laboratory medicine leaders need to understand which levers exist in their country – how healthcare programs are constructed and how value is ultimately recognized through payment models. That includes designing proactive healthcare, improving patient access, identifying risk, and, in the United States, engaging with quality and outcomes metrics that are embedded in the system.
If you know your system, you earn a seat at the table. Healthcare delivery is a team sport, and pathologists need to be able not only to participate, but to contribute meaningfully to the strategic design of healthcare programs. This isn’t optional. Pathologists have to know this stuff.
What does the shift toward predictive, data-driven care mean for the role of pathologists and laboratory medicine professionals in everyday healthcare delivery?
Ulysses Balis: It means placing laboratory medicine professionals and pathologists in roles that are much closer to how we traditionally think about primary care providers. We need to let go of that outdated notion of pathologists being hidden away in the basement, operating as a black box that nobody really understands. Laboratory medicine has a real opportunity to be front and center of both population-based care and care at the individual patient level.
Driving this change is an ability to generate predictive data that simply wasn’t possible even five years ago. That capability naturally forces new conversations with both clinicians and patients about how we incorporate this knowledge into routine care. There are several layers of exploration underway. First, what kinds of encoded data can we uncover that are truly proactive and predictive? And second, once we have that data, how do we use it effectively?
Frustratingly, it typically takes 11 to 20 years for new technologies and approaches to make it into clinical practice. We are on the verge of an embarrassment of riches – powerful analytical techniques and predictive tools – and we simply cannot afford a 20-year deployment cycle. One of Project Santa Fe’s core charges is to accelerate the translation of these discoveries, particularly in data analytics, so they can be applied for the greatest good as quickly as possible.
Khosrow Shotorbani: One of the key lessons we’ve learned is that reducing time to diagnosis – what we call diagnostic optimization – only creates value if it actually triggers a clinical protocol. That is where laboratories need a seat at the table. If we suspect sepsis, for example, that suspicion should automatically initiate a defined clinical response.
Instead, laboratories interface millions of results every day without knowing which of those data points lead to a clinical action and which do not. If pathology and clinical laboratory medicine are going to be relevant in the future state of healthcare, we have to be able to quantify that relationship – to understand and demonstrate which data trigger which clinical protocols.
How can laboratories move beyond that transactional testing model to ensure results actually drive clinical action?
UB: There is a tremendous opportunity to create closed-loop systems in which the laboratory plays an active role in ensuring results are acted upon appropriately. At Michigan, we’ve been doing this for about 15 years through one of the early computational pathology sections. For example, we perform thiopurine analog testing using machine learning rather than traditional bench-based assays, because it is more predictive. The analysis integrates molecular data with a complete blood count and comprehensive chemistry profile to generate a report for the clinician.
But it doesn’t stop there. We also receive feedback from the electronic health record that allows us to determine whether the clinician adjusted the patient’s immunosuppressive regimen in response to the findings – specifically whether azathioprine dosing is appropriate for inflammatory bowel disease.
We do similar things in anatomic pathology. Unexpected biopsy findings automatically trigger a requirement for the pathologist to contact the clinician and document the interaction. That repeats until confirmation is recorded, ensuring the clinician is aware of the result and that follow-up will occur.
What role do artificial intelligence and machine learning play in advancing the Clinical Lab 2.0 vision?
UB: When you step back and consider where much of medicine’s primary data actually resides, laboratory medicine and anatomic pathology represent a substantial portion of that foundation. What we are now recognizing is that the value is not just in the raw data itself, but the encoded, higher-dimensional information embedded within it – data that has existed all along, but has only recently become accessible to us.
Applying machine learning to enable predictive, rather than reactive, care management is enormously powerful, and pathology should be a central contributor to that effort.
At the same time, this is not a pathology-only endeavor. It requires participation from all groups that generate data as well as those positioned to act on it. From that perspective, Project Santa Fe sits right at the intersection of these needs. It represents an enabling layer that connects the dots: access to data, analytics expertise, data science teams, and the ability to deliver meaningful insights. In that sense, it’s a perfect storm – in a good way.
Can you give an example of how advanced analytics and integrated data could change patient care at both the individual and population level?
UB: One powerful example involves patients with rare diseases or unusual cancers, where there is often limited historical experience to guide treatment decisions. We can now integrate laboratory results with molecular, genomic, and broader clinical data, and compare them against large historical datasets with known outcomes. That allows us to build remarkably precise predictive tools.
Imagine being able to say, at the point of the initial biopsy, that a particular cancer will metastasize in six and a half months, with 98 percent predictive accuracy. The clinician could then explain to the patient that the analytics derived from their biopsy indicate an extremely aggressive tumor, justifying a more intensive treatment approach than standard protocols might suggest. That future represents precision, personalized medicine in its truest sense.
Pathology has traditionally been descriptive, answering the question, “What is this?” That role remains essential. But we are now poised to add another layer: understanding the biological potential of an individual disease. When you scale that insight across populations, the implications are even broader. For example, we know at a general level that microplastics are a concern, but are we truly tracking their biological impact across tens of thousands or millions of lives? With advanced data analytics, we could.
How should laboratories approach measuring their impact within value-based care models?
JC: For laboratories to engage meaningfully in value-based care, measurement has to begin at the outset. You need to know what you are looking for before you start. You can’t take action first and then try to figure out afterward whether it made a difference. The metrics have to be defined in advance.
Laboratories have to be able to prove that patients receiving care through them are better off because of that involvement. And that is actually very hard to do – but it is essential.
What is the most important mindset shift laboratorians need to make in a value-based care environment?
KS: Laboratorians need to understand who the customer actually is. It's not the person procuring the test result, but the entity at risk – financially and clinically – for outcomes. That includes those who influence or own policy.
From that perspective, value is defined by several factors. One is behavior: did the physician take action based on the laboratory’s recommendation, and did the patient comply with that action? Another is outcome. Clinically, outcomes are about early intervention and prevention. Financially, outcomes are reflected in optimization of measures such as key performance indicators, condition category adjustment, and the total cost of care delivery.
What skills gap does pathology need to address to remain relevant in value-based, population-focused care models?
JC: Thanks to leaders like Ulysses, and a global community within pathology informatics, a subset of pathologists and laboratory scientists now have training in data science. That represents one critical threshold we have to cross. The other, equally important threshold is understanding policy and payment.
My message to peers is this: not everyone in your practice group needs to master these areas, but someone does. The survival of your laboratory depends on having leadership that understands how value-based care, policy, and reimbursement actually work. That person does not have to be senior, but someone has to be in the big game, advancing the interests of the laboratory so it can function as the asset we’ve been describing.
I tell trainees the same thing. When you’re interviewing for jobs, ask whether your prospective employer understands these concepts. If a laboratory says, “Yes, we’re deeply involved, and here’s what we’re doing,” I would argue that laboratory is far more likely to succeed than one that responds with, “What are you talking about?”
How is pathology training evolving to prepare the next generation for data-driven, value-based care?
UB: Most programs now include informatics curricula and some exposure to management. What we still largely lack are formal components in areas like logistics and public policy. Those elements need to be added if we’re serious about preparing pathologists for the future state of healthcare.
JC: When I bring trainees to Project Santa Fe meetings, it genuinely opens their eyes. It often changes how they think about their careers. Seeing that shift is tremendously gratifying, because junior colleagues tend to embrace these ideas with real enthusiasm.
That said, once trainees graduate, reality sets in. They have demanding day jobs, heavy workloads, and often little or no protected time to engage in this kind of work. That’s why I come back to the same point: be part of a group where someone is doing this. As a more junior person, bring your talents to that effort. Engaging in this work can be a real career accelerator.
How can laboratories sustain and fund strategic, value-based activities ?
JC: This challenge has followed the movement from the very beginning. In the United States, there is typically no direct payment for these kinds of strategic activities. At Northwell, it ultimately comes down to the annual budget. We bring forward our strategic initiatives, along with their projected impact on total system profit and loss. If the budget is approved, we move ahead. Under that approved budget, our responsibility is to outperform it.
For independent laboratories, however, this is much more difficult. That’s why I often point to TriCore Laboratories in Albuquerque, New Mexico, which has demonstrated that it can be done through separately negotiated funds flow to support value-added activities.
A third pathway is exemplified by the University of Michigan, which has successfully obtained payment codes for predictive analytics.
What role can pathologists play in direct patient engagement, and how might that change the future of care delivery?
UB: Jeff Myers started a pilot program several years ago that has been remarkably successful. In this model, patients who have been diagnosed with cancer can schedule dedicated time to meet one-on-one with the pathologist who signed out their case. They review the histology together in an interactive setting.
For patients, this has been incredibly informative and, in many cases, cathartic. It helps prepare them emotionally and intellectually for the therapeutic journey ahead. Patients describe these sessions as transformative, giving them an anatomic frame of reference for their disease and helping them feel ready for what comes next. Many have expressed deep gratitude for the experience.
JC: I’ve done this myself on an ad hoc basis, and one patient captured it perfectly when they said, “Now I know what I’m fighting.” That kind of clarity can be incredibly powerful.
KS: I would add that this concept is not entirely new. In the United States and Europe, we are actually behind in regard to patient engagement. In a recent publication, we argued that a future world-class clinical laboratory does not start with science alone. It starts with knowing where patients are.
The pandemic reinforced that lesson. We can no longer build “Taj Mahals” of healthcare and wait for sick patients to arrive. Patient engagement, in my view, is the next blockbuster drug of the century. Patients are not passive recipients of care; they are members of the care team. We need to do things with them, not to them.
Newsletters
Receive the latest pathologist news, personalities, education, and career development – weekly to your inbox.
