The Radiologist and Population Health Management

Far from being marginalized, radiology can survive and thrive as one of the most valuable players in a population health world, but it will need to prove its strategic value to health systems.

The concept of population health management is often spoken of as a solution to the perverse incentives inherent in fee-for-service payment that encourage turf battles between specialties and quantity of care over quality of care. As much as radiologists will agree on the need for more appropriate utilization of imaging services, greater efficiency and more collegiality between specialties, the idea is also anxiety inducing.

The reason radiologists worry about population health is a good one. When the concept of managing the health of a population of patients is linked to a payment model such as the global payment or capitated model, it flips the traditional radiology business model upside down. Radiology procedures, tests and related services that used to be revenue centers become cost centers within the organization.

Nonetheless, the specialty can survive and even thrive on its strategic value rather than being a direct source of profits, say experts Radiology Business Journal consulted for this article. It will, however, require a new and possibly 180-degree-different practice model.

Radiology as strategic cost

One population health expert who is confident about radiology’s ability to adapt to population health is Paul J. Chang, MD, FSIIM, professor and vice chair, radiology informatics, University of Chicago School of Medicine, as well as medical director of both enterprise imaging and service-oriented architecture (SOA) infrastructure at the University of Chicago Hospitals.

“Radiology can thrive and should thrive as a cost center,” he says. “This fear that people talk about of moving from a revenue center to a cost center, which will happen, is actually an opportunity. It just means that we will have to change. IT is the perfect example. IT has been a cost center from day one. Who has the biggest budget? IT. Our department has one of the biggest budgets in the hospital these days, and we pay our people very, very well, because we are strategic. We are critical to the mission.”

In his view, there are four main ways radiology can participate in population health and survive on its strategic value to the enterprise.

  • Advising other doctors through a range of consultative services and acting as a gatekeeper on healthcare services.
  • Assuming responsibility for disease surveillance.
  • Directing decision support systems, including overseeing the implementation of clinical decision support (CDS) tools as part of computerized physician order entry (CPOE).
  • Applying radiology’s background in analytics to the enterprise level of an organization

However, at this point, relatively little of this is being done, Chang says, because the U.S. healthcare reimbursement system is in a gap period where even though there is a lot of talk about population health and possible new reimbursement models, the primary payment model is still fee-for-service.

“People tend to be very rational and they are not going to pull the trigger on a change in world view unless it is necessary to do so,” he says. “When people talk about the declining role of radiology in population health, my response is ‘stay tuned.’ This is all still, in many ways, our future.”

Living in a fee-for-service world

Work already is underway at forward-thinking institutions to improve care efficiency, quality and safety, as providers strive to meet quality incentive payments in the fee-for-service world and begin to take on capitated contracts and global payments in a population health world.

The University of Chicago, for example, is building a database of incidental findings recovered from radiology reports, of which only 30% are being followed up on, Chang says. If a recommended follow-up test or procedure is not ordered, the application can escalate matters all the way up to sending a certified letter directly to the patient about the finding and next recommended steps in care.

This application is part of an overall enterprise infrastructure initiative at the University of Chicago called Enterprise Informatics and Analytics. The same infrastructure also supports breast imaging communication and escalation.  Chang notes that this may be a model for how radiology will someday take ownership of disease surveillance in a population of patients instead of simply trusting that other physicians will read their reports and take the appropriate actions.

Another example can be found at Dartmouth-Hitchcock Health System in Lebanon, N.H. Under the leadership of neuroradiologist Clifford J. Belden, MD, the health system focused on creating standards of care that took the value of imaging services into account and implementing decision support tools. It also improved its ability to import scans done at other institutions into its PACS to avoid repeating imaging and worked with other regional healthcare systems on refining trauma imaging protocols, so that when a trauma patient is transferred to Dartmouth-Hitchcock Health System, studies done elsewhere don’t need to be repeated because the original study was done incorrectly.

Belden was recently appointed Dartmouth-Hitchcock Health System’s chief clinical officer and executive vice president, integrated delivery system, part of a leadership team that will implement Dartmouth-Hitchcock’s new operating strategy. He points out that even though they’ve been able to reduce imaging utilization to one of the lowest rates among Medicare providers, radiology is still a profit center for the health system and payment arrangements are overwhelmingly fee-for-service. Their efforts are about doing what is right and laying the groundwork for the future.

“At Dartmouth, we are well over a decade into this change, so it is not something that is particularly new for us, but it does present some opportunities,” Belden says. “The goal of the Affordable Care Act, if you want to think about it that broadly, is really around improving the cost of care and quality of care, and radiology has been a fixture in helping drive improvements in cost and quality. Twenty years ago, one of the most common admissions to the hospital was for exploratory laparotomy to see what is going on in the belly or admissions for observation for abdominal pain. We don’t do those things anymore because imaging has really impacted the health and well being of our patients in very positive ways. Imaging is a part of the high quality, low cost provision of care.”

A model to watch

Where a true population health payment model is brewing is on the other side of the country. Anthem Blue Cross has partnered with seven Los Angeles and Orange County health systems to create Vivity, a health insurance product that pays the systems a set fee for each individual’s care. If the systems can care for each Vivity member’s health for less than the set fee, they keep the difference. If they go over, they absorb the cost.

Cedars-Sinai Medical Center in Los Angeles is part of Vivity, and Barry D. Pressman, MD, FACR, professor and chair, department of imaging, has been working hard to prepare his practice at the Cedars-Sinai S. Mark Taper Foundation Imaging Center for the type of capitated-payment population health model that Vivity may turn out to be. (The exact payment arrangements for the Vivity-participating health systems are still being worked out.)

“It is the worry about the cost that is driving everything right now,” he says. It’s Pressman’s belief that population health on its own does not need to be tied to a particular payment system in order to curb inappropriate imaging use while simultaneously keeping patients healthier with better screening and disease surveillance.

“We are trying to help the hospital on the cost side, using our involvement with them and our willingness to participate to help demonstrate that we care and that we are believable, so that when we say the study should be done, that is just as valuable as when we say the study shouldn’t be done,” he adds.

Pressman identifies the following ten steps that radiology needs to take to prepare for the transition to population health and capitation.

  • Decrease inappropriate and harmful testing.
  • Harness more big data to determine the true costs of delayed diagnostic imaging.
  • Band together in ways that do not violate anti-trust laws to ensure a voice in health system strategic planning.
  • Become involved in health system leadership and encouraging more radiologists to also earn masters in health system administration.
  • Move to structured reporting and standardized terminology to make reports easier to read and use by patients, clinicians and researchers.
  • Lead the implementation of clinical decision support and catching orders that “don’t make sense.”
  • Be involved in setting clinical guidelines.
  • Push for more preventive techniques like mammography that allow them to manage the health of a population (e.g., low-dose CT lung cancer screening and virtual colonoscopy).
  • Set the clinical protocols.
  • Relocate to where the clinicians are, either with an onsite presence or a virtual one, to enable more collaboration on studies.

“We are trying to put some of our reading rooms back where the clinicians are,” Pressman says. “For example, we are putting a musculoskeletal reading room into the new spine center where the surgeons are so that there is more interaction. That leads to faster decision making and faster treatment, which in turn reduces costs.”

Return of the ‘doctor’s doctor’

One of the upsides of the trend toward population health management is that it should elevate radiologists’ consultative role. “I always feel I’m providing the most value for the patient when I’m sitting in a tumor board with the treating physicians and identifying where the tumor is and participating in the care planning. That is not reimbursed,” Belden says.

“Talking with a patient isn’t reimbursed,” he adds. “Answering a patient’s questions isn’t reimbursed. Getting on the phone with an ordering physician and telling him or her that a CT isn’t a good test for this patient is not reimbursed.  That is part of the benefit of moving to a new model: Your reimbursement doesn’t hinge on how many CT reports you sign today.”

Chang was instrumental in the development of one of the early web-based PACS and notes that this technology really changed medicine from an era where rounds started in the radiology department—because that is where the film was—to an era where images could be viewed anywhere. The radiologist doesn’t even need to be onsite.

“Back when we were purely the revenue generator, we didn’t have to deal with anyone else and we just needed to worry about how to be the most efficient,” Chang says. “We isolated ourselves, and we need to go back to the days when no one questioned our strategic importance because we were the doctor’s doctor. No one read our reports back then because they didn’t have to. We collaborated many times a day instead of just sending reports that are like messages in a bottle because they may or may not be read.”

Radiologists as the ‘honest brokers’

In some ways, fee-for-service systems that made radiology a revenue generator minimized the radiologist’s more important role as the intermediary that ensured that care design and funds align to reward all clinicians for delivering higher-value care. In an opinion piece in the July issue of the Journal of the American Medical Association, Steven Seltzer, MD, chair, department of radiology, Brigham and Women’s Hospital and Harvard Medical School, emphasizes this need for the “honest broker” who can partner with integrated delivery systems to align doing the right thing for every patient with also doing the right thing for the system.

In Seltzer’s view, radiologists must partner with integrated delivery systems to manage utilization and costs in four ways.

  • Overseeing the content and implementation of CDS.
  • Performing a range of consulting services.
  • Tracking the appropriate use of imaging tests at an individual clinical level.
  • Leveraging lower-cost community sites to provide the right care, at the right time and in the right setting.

They also need to better promote image-guided interventions as a low cost but high quality alternative to more costly procedures, something Seltzer noted the field has not always been very good at. “We’ve not been very good at taking credit for our good work,” he says.

Most importantly, Seltzer believes that radiologists need to stop saying they are just readers and become true consultants working with all clinicians, from the primary care nurse practitioner to the head of oncologic surgery. Chang concurs.

Chang confesses that like most radiologists, he has not always been quick to take calls from primary care because reports are written for experts like surgeons who are morphologically driven, and he knows that a call from primary care will mean a long conversation that goes into care planning and next steps. In population-health world, the payment model incentive to get off the phone and get back to reading images as quickly as possible goes away. What has value in this new model is the conversation that leads to a more appropriate use of health resources and a quicker recovery for the patient.

“Shame on me that I didn’t want to answer that call because it was going to be a long conversation,” Chang says. “In a population health world, we want to encourage those types of discussions, but perhaps in a more efficient way than a phone call.”

The embedded radiologist

Giving radiology a virtual presence in primary care is already happening at Brigham and Women’s Hospital, where the health system is testing a program that assigns radiologists and radiology extenders to be telephonically and electronically present in four patient-centered medical home primary care practices. “We are saying, if you have a question about whether a patient should have a CT scan or an MRI, helping you make that decision is part of our job and we are part of your team,” Seltzer says.

In addition, when the Dana Farber Cancer Institute built a new outpatient center, Seltzer’s department asked to have clinical space at the center specifically to facilitate consulting. “This was getting away from the traditional idea that I grew up with that all the radiologists should be in a room in the basement because that was close to the film library,” Seltzer says. “Now that we are untethered from the acquisition devices because we can move the images electronically wherever we want and untethered from the film library, we are free to ask, where do we want to be and do we want to be in the clinics?”

The model to follow, Seltzer adds, is breast imaging, because it is the part of the field of radiology that is already engaged in population health management through disease screening and communicating directly with patients. He is not alone in that opinion.

“When I design systems for the University of Chicago, I always say that every one of us in radiology needs to be like breast imagers,” Chang says. “Breast imaging is a great model that shows us the future. Breast imagers are considered critical, strategic parts of the care management team with the surgeon, the oncologist and the pathologist. Breast imagers’ importance to and direct communication with the consumer is a given. We don’t care if the ordering physician reads our mammography report because we communicate directly to the health consumer.”

CT lung cancer screening and CT colon cancer screening are prime targets where the mammography model of population health management could be applied. “Within a few years, there may be a portfolio of situations where radiologists are the population health managers,” Seltzer says. “That is a role that we can embrace because we’ve actually been doing it for decades with mammography. We have just never called it population health.”

Where population health could go wrong

Is the fear of population health management unfounded? No.  

The danger in moving to a population–health-management model, our experts say, is that if radiology is not at the decision table to demonstrate its strategic value, radiologists may simply see all imaging use cut back with no new compensation for their consultative services and possibly harmful implications for patient outcomes.

“If you take it to an extreme where every service you offer cuts into your profitability, you could say, ‘Well, I just won’t provide any services,’ and keep the money,” Seltzer says. “That can’t happen.”

Radiology must be at the table with as much analytical data as possible that validates the sepcialty’s contribution, both in diagnostic imaging and in the development of newer image-guided interventions that create low-cost alternatives to more expensive treatments, he explains.

Belden adds this: “There is a risk, if we are not at the table, that our value is still measured in RVUs and the number of interpretations. In that case, when you put in a system that puts in some type of regulator on how many of those tests are done, there is no way to capture dollars for the additional activities that you are doing.”

Radiologists also need to ensure that the system is not rigged in such a way that ordering physicians hesitate to use imaging out of fear that they will look like they are over-utilizing testing, even though all of the tests they ordered may be appropriate for their patients and their health system environment.

“We have to recognize that there is underutilization [as well as overutilization], which results in diseases being detected in more advanced states,” Belden says. “Another form of underutilization is when we consult with specialists rather than doing imaging from the primary care perspective, or don’t use imaging that can help decrease the length of stay, such as when evaluating stroke symptoms.”

Radiologists in a population health model linked to a payment system that incentivizes controlling costs have a dual role in both ensuring the broader use of technologies that can save lives and lower costs, and limiting the use of expensive tests that add relatively little value in terms of patient outcomes. Making sure that the right balance between these two contradictory goals is set at the enterprise level of an organization is something in which radiologists must be involved.

“You have to be at the table,” says Pressman. “If you are asking what’s for dinner, it’s you. The best way to be involved is not at the national level, although that is important, but by responding to, as well as directing, what is going on within your institution or in your environment.”

Getting involved at the enterprise level

Pressman says that although even the ACOs Cedars is currently participating in are fee-for-service, he is watching the evolution of payment models closely and is constantly communicating with Cedars-Sinai leadership about their two ACOs. He also is thinking ahead to how his practice would operate in an environment where payment is capitated.

That means helping the hospital evaluate every possible option, not just CDS, but also outside radiology benefit management (RBM) services and possibly having the radiology department act as an RBM. “In a capitated model, it is to our advantage to reduce inappropriate imaging because our reimbursement is capitated as well,” he notes.

His practice has been intimately involved with helping Cedars-Sinai improve the appropriate ordering of imaging studies. The practice created significant patient safety improvements by implementing a system that forces the ordering physicians to do an override when ordering a test that does not follow one of the more than 60 Choosing Wisely appropriate use criteria that apply to imaging.

It also sends the physicians reports on the appropriateness of their imaging ordering so that they can improve. In addition, Pressman says he has personally brought in the CDS vendors being considered for the addition of CDS in the hospital’s Epic electronic medical record system.

At Dartmouth, the radiologists help collaborate on its “Knowledge Map” system for setting best practices and developing decision support tools. “There is a tendency to think of radiology decision support as different from other forms of decision support, and in some ways, it is more mature because we have the ACR Select product and have been doing this work for some time,” Belden says.

“But if you are thinking about how to do decision support within your institution, this should be within the framework of how you will do the rest of decision support within the institution,” he continues. “It shouldn’t be just radiology. There are as many opportunities for savings within lab, pharmacy and specialist referring and care. “

What makes radiology so strategic to healthcare systems contemplating population health is that although the field may have withdrawn from its more collaborative role and become somewhat commoditized under fee-for-service, at its heart, it is integral to disease screening, disease surveillance, collaborative care planning and overall cost control. The upside of population health is that it could return radiology to an earlier era where it was at the strategic center of healthcare.

“Radiologists have a unique lens through which to look at the health system because they interact with a lot of the specialty and primary care physicians,” Belden explains. “There is a level of trust and a level of understanding of disease processes that anyone who has worked in a busy imaging environment knows. … Radiologists are in a position to add a lot of value to the system, but they have to get engaged.  It is not just controlling imaging, but helping define the best ways to evaluate patients.”

Lena Kauffman is a contributing writer for Radiology Business Journal.