From Value-Based Care to AI, Imaging Leaders Look to Radiology’s Future

Oct. 10, 2016
Medical imaging leaders face a number of challenges in the ongoing transformation of care delivery, yet many of the key market forces that are changing the field also are pushing imaging toward innovation, according to imaging informatics leaders.

Medical imaging leaders are facing a number of challenges in the ongoing transformation of care delivery, yet many of the key market forces that are changing the field also are pushing imaging toward innovation, according to many imaging informatics leaders at a recent conference in New York City.

At a conference focused on driving innovation in imaging, leading radiologists and imaging informaticists shared their perspectives on the future of medical imaging and the role that imaging plays in the transition to value-based care and population health initiatives. The event was sponsored by New York City-based Ambra Health, formerly DICOM Grid, a medical data and image management company.

There are key industry forces pushing radiology to innovate, notably care delivery transformation, imaging consumerism and the overall growth outlook for the radiology field, Lea Halim, senior consultant at Washington, D.C.-based The Advisory Board’s Research and Insights division, said.

Halim also said there are industry and economic trends impacting the growth outlook for imaging. The diagnostic imaging services market is saturated, as rapid expansion in the provider landscape has quickly absorbed available market share and, in recent years, there have been fewer new modality or advanced technology opportunities, and, at the same time, economy-driven influences are driving patients to put off radiology exams, she said. “Imaging leaders need to look at new growth opportunities alongside traditional outpatient growth,” she said.

Halim said imaging leaders need to focus on and invest in expanding screening exam services, as screening provides a unique opportunity for imaging’s involvement in the shift to value-based payment and care delivery. An aging population combined with an emphasis on population health initiatives means the demand for screening exam services will likely grow, she said, and imaging leaders should consider leveraging screening programs to promote early diagnosis within their patient populations. Beyond mammography screening exams, imaging leaders should consider providing lung cancer screening, CT colonography, abdominal aortic aneurysm (AAA) screening and even the more exploratory cardiac CT for calcium scoring screening.

With more providers taking on risk through participation in accountable care organizations (ACOs) and mandatory bundled payment models, radiology’s hospital partners and referring providers are increasingly focused on controlling total cost and utilization management.

“There is a view is that care delivery transformation has had little effect on imaging—you’re not directly taking on risk or involved in new payment models. But, you don’t get a free pass. It’s relevant to radiologists and radiology providers. With the proliferation of alternative payment models it impacts how providers think about cost as more are taking on risk,” she said.

The simple truth, she said, is that diagnostic imaging is emerging as a top area in which there is an opportunity for cost savings. However, it also could be an opportunity for imaging to be a care transformation partner by helping provides share images as well as standardizing protocols and standardizing care.

“Imaging must establish its value under population health management, as the role of imaging is not as clear as we’d like it to be. If you’re not at the table, then you’re on the menu as far as utilization reduction when it comes to population health strategy,” she said.

According to Halim, there are three key opportunities for imaging leaders to align imaging initiatives with health systems’ population health goals—utilization management, which can improve quality and reduce costs; screening programs, which can impact patient risk escalation; and incidental finding management, which can help keep patients loyal to the health system.

“Moving forward, how do I communicate my value to the CEO? How do I communicate what imaging contributes to population health effort? How do we communicate that we’re the people who hold the key to appropriate diagnosis and treatment?” Halim asked.  

What Does “Quality” Look Like for Radiology

Measuring and demonstrating quality in the radiology field in the ongoing shift to value-based care is a substantial challenge, according to imaging informatics leaders at the conference.

Geraldine McGinty, M.D., a radiologist and assistant chief contracting officer at New York City-based Weill Cornell Medicine, said, “Most of the population health work that I’ve been involved in has been around primary care and its important work to manage chronic disease. We have a lot to contribute and we need to develop better ways to demonstrate how we contribute and measure how we are contributing.”

She continued, “With the MACRA [Medicare Access and CHIP Reauthorization Act of 2015] legislation, there have been some specialty-specific metrics and there has been some work crafting episodes of care that are imaging focused. We have a huge opportunity to demonstrate our value and develop metrics so we will be rewarded for that.”

“What impact does radiology have? How does what I do impact care and affect the lives of patients? The challenge is that we really don’t know,” Eliot Siegel, M.D., professor and vice chair, research informatics systems, at the University of Maryland School of Medicine, department of diagnostic radiology, said.

“There have been discussions about studying outcomes. Is the radiology helping my patients? And looking at appropriateness criteria. One area you can measure is the outcomes and the definition of quality has to be tied to outcomes, but it’s a big black box right now. I’d like to see more efforts track the impact it has on patients,” said Siegel, who also is chief of radiology and nuclear medicine for the Veterans Affairs Maryland Healthcare System.

A CT scan performed at the right time can have a significant impact on patient outcomes, “but the question is, what is our actual cost to providing care and contrast that to the impact on care?” Siegel said.

Halim outlined several ways imaging can succeed under new care delivery and payment models. Under an ACO model, imaging leaders can provide imaging sharing technology to ACO referring provider partners and leverage screening programs to promote early diagnosis. When hospital partners and referring providers are participating in a bundled payment model, imaging leaders can identify utilization trends of bundled services and leverage freestanding sites to minimize imaging follow-up costs, she said. With a capitated payment model, radiologists can promote appropriate ordering to reduce unnecessary imaging and create incidental finding protocol to ensure necessary follow-up care, she said.

“I think image sharing should be an area of extensive focus for imaging leaders. When they take on the responsibility for the health of a population, the ability to view images in real time and to share images throughout the course of a patient’s treatment can be critical to keeping down costs and elevating quality,” she said.

The Impact of Machine Learning and AI

McGinty and Siegel both see machine learning and artificial intelligence having a substantial impact on radiology.

Siegel cited a recent article published in the Journal of the American College of Radiology titled, “The End of Radiology? Three Threats to the Future of Practice of Radiology,” which cautioned that machine intelligence could end radiology as a thriving specialty.

“I can reassure you that there is no chance that radiologists will be replaced by computers,” he said. “We’ve awakened to realize that the technology is out there and there are some very interesting applications.”

According to The Advisory Board, machine learning is a set of algorithms that can learn complex patterns and make predictions from data. Deep learning is a subset of machine learning that aims to mimic the way the human brain functions by attempting to create artificial neural networks. This type of learning has shown a lot of promise in areas important to radiology, such as image recognition.

“While an algorithm can correctly identify a picture of a dog, reading an MRI or CT or doing other modalities is a different challenge,” Siegel said. “At this point, so far, they are just barely scratching the surface.”

“It’s something we have to really think about and I’m engaged and excited about it and it has the potential to expand the scope of what we can do. We have nothing to gain by sticking our heads in the sand,” McGinty said.

Looking ahead, there is the potential for deep learning algorithms to transform the radiologist’s role. Algorithms that pre-read images for radiologists could vastly cut down on read times, improve radiologist accuracy, and shift some workload from reading to reporting.

Imaging Consumerism

Halim of The Advisory Board sees the rise of healthcare consumerism as having a long-term impact on imaging.

“Consumers are increasingly shopping for healthcare, and imaging leads the pack for services that are shoppable, so we must deliver on their preferences,” she said. Citing on-demand services such as Uber, she noted, “Where consumers are coming from, they are used to getting what we want, when we want it, at a price that’s acceptable.”

Imaging providers need to adapt to consumers demands by providing transparent prices which entails providing price estimates to patients and offering price competitive options. Imaging providers also need to provide timely care by extending their hours, offering online scheduling and providing timely results. “We need to offer convenient locations, such as having accessible freestanding clinics, and billing on a lower fee schedule,” she said.