From Volume to Value: Radiology’s New IT Value Platform

“Radiologists have been talking about the value paradigm for a number of years,” says Jonathan B. Kruskal, MD, PhD, FACR, chairman, department of radiology and radiologist-in-chief, Beth Israel Deaconess Medical Center and chair of the American College of Radiology’s Quality Management Committee. “It is way past due that all radiologists embrace every component of this new reality of our world.” 

According to Kruskal, the elements of value that a service provides are broken down into the value equation, which links quality, appropriateness and outcomes “inversely to cost.” In practical terms, radiologists need to approach the way that they manage measurements obtained when evaluating the quality of services that they provide. “This also means that we all must embrace appropriateness since an inappropriate study that we perform essentially is of absolutely no value to any stakeholder,” Kruskal emphasizes.

When it comes to issues regarding appropriateness and outcomes, radiology information technology does, and will continue to play a crucial role. IT can be used to demonstrate utilization and outcomes, and by extension appropriateness, says Jon Copland, Inland Imaging, Wash., CIO and CEO of the practice’s information technology services subsidiary.

As an example, he points out that it’s not unusual for primary care providers to utilize chest X-rays to see if a patient has pneumonia or tuberculosis, and it should be easy to track how often the imaging results in a negative result. If the percentage is relatively high, then there’s an appropriateness issue.

At the same time, Copland says, one can show that an eye, ear and nose physician who sends a large number of patients for a CT exam is a high utilizer of imaging, but that the number of exams that come up positive could be relatively high.

“We can do a statistical analysis of appropriateness of care,” Copland says, adding that it’s also a way of demonstrating value in a value-based healthcare model, as opposed to the traditional fee-for-service model in which “you get paid for services, whether they’re appropriate or not.”

This transition to a value-based model is one that organized radiology has long been preparing for, Kruskal says. “Organized radiology has taken many steps over many years to help practices prepare for the value-based payment schemes,” he says. “In particular, the American College of Radiology has played a yeoman’s role in not only making every effort to spread the concept and challenges and opportunities inherent in this paradigm through its Imaging 3.0 program, but they’ve also strived to develop and deploy a host of tools that can be utilized to measure quality, appropriateness and outcomes.” One of those tools is clinical decision support.

Clinical decision support

The concept of clinical decision support (CDS) has been around for quite a long time. For example, Massachusetts General Hospital—in the absence of readily available commercial computerized order-entry systems and CDS tools—developed its own decision support system in 2004 for advanced imaging examinations such as CT, MRI, nuclear medicine and PET.

The passage of the American Reinvestment and Recovery Act, and the accompanying Health Information Technology for Economic and Clinical Health Act, and its support for the concept of electronic health records, led to the broad implementation of EMRs, which, in turn, has created an environment favorable to the development and deployment of CDS.

“Hundreds of hospitals have de- ployed the technology,” says Keith Dreyer, DO, PhD, FACR, FSIIM, vice chair of radiology, and director, Center for Clinical Data Science, Massachusetts General Hospital, Boston. “It’s integrated into most all of the major electronic health records.”

With the fairly rapid deployment of EHRs into American health systems, and the fact that CDS has been shown to reduce inappropriate imaging—and by extension improve patient care and cut costs—a provision of the Protecting Access to Medicare Act (PAMA) requires physicians to use CDS tools and document their use whenever they order advanced imaging tests.

PAMA, which was signed into law by President Obama in April 2014, mandates that physicians utilize appropriate use criteria through CDS when ordering advanced imaging studies such as diagnostic MRI, CT, and nuclear medicine. It also only applies to outpatient settings such as physician offices, ambulatory surgical centers and hospital outpatient departments. Originally set to debut in 2017, CMS has delayed implementation until January of 2018 and is working on the second of four rounds of rulemaking (see sidebar, page 00).

While there’s been much talk about clinical decision support over the years as a way to manage imaging ordering appropriateness, the passage of PAMA has created more urgency when it comes to implementing CDS. Its arrival, however, raises concern and the following questions: How will CDS affect individual radiology practices and their referring physicians? More specifically, how will it affect the consultative relationship between radiologist and provider?

A collaborative approach

One program designed to try to deal with those issues is called R-SCAN, the Radiology Support, Communication, and Alignment Network. R-SCAN was born out of a CMS grant under Medicare’s Transforming Clinical Practice Initiative with the goal of helping radiologists and referring clinicians to collaborate more effectively in order to improve imaging appropriateness

R-SCAN is closely aligned with ACR’s Imaging 3.0 initiative. It is best described as a collaborative action plan that brings radiologists and referring clinicians together to improve imaging appropriateness based upon an ever-growing list of Choosing Wisely topics.

The program is web-based and free for both radiologists and clinicians. It includes a step-by-step guide that leads participants through the process of ordering appropriate imaging exams based on recommendations in chest, cardiac, trauma, genitourinary, neuro and musculoskeletal imaging. Participants use a customized version of ACR Select, a computer-based diagnostic imaging decision-support system that deploys ACR Appropriateness Critiera® to rate imaging exams ordered, improve imaging appropriateness and create project reports that can be used to implement and quantify changes in the order process.

“The beauty of it is that it is really made to be a one-stop shopping resource for someone who doesn’t have a lot of experience in quality improvement,” says Marc Willis, DO, associate professor of radiology, Baylor College of Medicine, Houston, Tex., and a clinical advisor on R-SCAN. “It takes them step-by-step through a quality improvement project and it’s online, so they easily can input the data and get the feedback needed.”

Its purpose is to promote collaboration between radiologists and clinical referring providers to address the issue of imaging appropriateness, Willis says. “It’s also really allows radiologists to lead that effort,” he adds.

One way to think of R-SCAN, Willis says, is that it’s an excellent first step towards implementing CDS. With the PAMA mandate and the knowledge that as of January 2018 there will be legal requirements to use clinical decision support, providers have to move quickly.

“Given the budgeting cycles that most healthcare systems and hospitals have to go through to implement any new technology, it’s a good time to start getting things headed in the right direction,” Willis says. “It’s a great way for radiologists to take on this issue, to work with their referring providers and start getting people exposed to clinical decision support.”

IT as measuring stick

Measuring value in radiology has challenges beyond the scope of implementing clinical decision support, Willis believes. He thinks that radiology practices “need to start looking at creative ways in which we measure and market the value that we do provide.”

While it is important to look at issues related to imaging appropriateness because it’s a demonstrable way of reducing healthcare costs, Willis adds that when it comes to adding value, “We simultaneously should be looking at value-based metrics, which really comes down to patient outcomes.”

The challenge, he says, is that most of the metrics used in healthcare are process metrics. “That’s a difficult space,” Willis says. “We have all of this information, but we really don’t have mature mechanisms to get meaningful data out. That’s a great need right now, because the more we start thinking about measuring those patient value-based outcomes, the better off we’ll be.”

Radiology definitely has the IT tools to demonstrate value, Kruskal believes, but radiology practices and departments need to understand value and to utilize the tools appropriately. “There are more than enough tools for measuring quality, appropriateness and outcomes, but what is required is that these metrics be collected, carefully analyzed and managed so that we can change the balance and provide recognized value,” he says.

A big challenge facing radiologists is deciding which specific performance metrics should be utilized. “Many people have been paralyzed by the inability to decide which metrics to collect and manage,” he notes. “This analysis paralysis has been to our detriment, as we look at our physician colleagues in other services who have been collecting and managing quality metrics for a number of years,” Kruskal says.

“It does not necessarily matter which specific metrics we choose since continuous performance improvement is a journey comprising innumerable, small, incremental steps,” he continues. “What would help tremendously is the availability of national performance benchmarks which would then allow practices to know where they stand compared to their colleagues, and it would also allow us as a profession to introduce efforts at all levels to improve the quality of services that we provide.”

Willis makes a similar point. “We should think about having some kind of basic portfolio of standardized patient-based value outcomes that we can start to measure,” he says, “If we are all measuring different metrics, what exactly will that mean? You’ll be able to benchmark internally, but you’ll never be able to benchmark externally in a meaningful way.”

Kruskal agrees. Without benchmarks, he adds, radiologists will have no effective way of marketing and branding the service that they provide.

Generating transactions to adding value

Raym Geis, MD, clinical assistant professor, department of radiology, University of Colorado, Boulder, is optimistic about radiology’s ability to demonstrate its value. “This is very scary time for radiology,” he says. “Historically hospitals and healthcare systems felt that radiologists were

valuable because they knew that radiology departments helped bring in a lot of money. Now, as you shift to some sort of a PPO, or bundled or fixed-payment model, that equation no longer works.”

When Geis talks asks hospital CEOs what makes radiologists valuable and hopw they determine that value, he says he invariably hears: “I don’t know how to value my radiologists.” Geis recounted a conversation with one CEO who referred to radiology as a big black box in which a hospital could invest a little bit of money at the top of the box, and see much more come out of the bottom.

He figured radiology was valuable, the CEO told Geis. Now that the system has changed, there’s not as much money coming out of the bottom.

Looking at it that way, radiology doesn’t appear to have much value, Geis says. And while hospital executives like the one Geis quoted realize that’s not true, “They don’t know how to value it,” he says. “That’s radiology’s big issue right now.”

Geis believes radiology has to be able to find the right kind of data regarding the contributions it makes to providing better patient care, better quality and improved outcomes, and convert that into dollars and cents so that it can demonstrated it value. He believes informatics is the anwer.

“We are going to have to rely on machines to help us figure out whether we are doing a good job and demonstrate the value we provide all over the healthcare system,” says Geis, who is also vice chair of ACR’s Informatics Commission. “It’s going to require new computer software algorithms.”

According to Geis, these algorithms will be used to look for patterns in very large and complicated amounts of data. These patterns will help radiologists and healthcare system executives see how radiology is performing when it comes to providing better quality and improving outcomes.

“We need those computer programs to figure it out, and they haven’t been written yet,” Geis says. “The IT ability is there. We just don’t have the tools yet.”

Having said that, Geis believes radiology is well placed since it is implementing the IT tools—such as speech recognition software and structured reporting—needed to accumulate data. “We have data that is machine readable, and that has always been a limited factor in figuring these things out,” he says. He points out that when IBM Watson outperformed its human challengers on the TV game show Jeopardy in 2011, it was using algorithms that had been written years earlier. It was Watson’s ability to read an immense amount of data—200 million pages worth, including the full text of Wikipedia—that made it all work.

“Radiologists are ahead of the game in terms of data, but we do need algorithms to answer our questions,” Geis says. “I’m optimistic radiology can do it, since we are focused on demonstrating value. That’s not something you hear being talked about in other specialties on a daily basis.”

Costs and reimbursement

Providing high quality radiology that deomonstrably delivers quality will have significant costs. “The costs of good quality are high and do include the ability to provide timely informatics tools and do typically require personnel, including nursing staff, project managers, data analysts,” says Kruskal. “The staff required depends largely on the practice profile and size.” 

“ In our experience,” he continues, “one cannot rely on nationally available software tools to spit out data and assume that the data will be correctly analyzed and managed.  The truth is that the costs of providing poor quality service are far greater than the costs of providing good service, and we are all well aware of practices that have succumbed to providing poor quality.”

Copland points out that radiology has implemented excellent IT solutions when it comes to processes like ordering, decision support and patient-tracking. On the other hand, he says, “there has been this chasm between clinical care systems and the financial medical billing system, which we are trying to bridge.”

According to Copeland, Inland Imaging recently hired a data analyst for just that purpose. “He’ll be pouring through the millions of exams we have, looking at financial outcomes and reimbursement, and trying to tie together the clinical and financial components,” Copeland says. “If we get declined by the payor for doing an MRI, we want to have systems up front that will tell us that the payor is not going to pay for this.

“We’re pretty good at that now, based on the pre-authorizations that are required. But there is room for improvement between the clinical, financial and insurance payor systems, which are somewhat independent of each other, even though we do our best to tie them together.”

Dreyer notes that the issue of compensation is related to clinical decision support and is one reason why radiologists so readily support CDS. “There has been this strong drive [on the part of the ACR] to put decision support in front of ordering physicians, because overutilization often comes back to the radiologist,” he says. “The federal government is not going to just keep paying more for unnecessary exams. The challenge is that radiologists themselves don’t order those exams, nor can they stop the ordering of those exams.”

Most radiologists support this aspect of decision support because it helps to manage the overutilization for which they are penalized, but can’t control. However, there is an aspect of the CDS mandate that will present a challenge when it comes to compensation, and that’s how the claims process will work.

“Obviously, for payment to occur we will have to submit a claim,” Dreyer says. “In advanced diagnostic imaging, we also have to submit some sort of proof that the order physician consulted appropriate use criteria.”

The problem is that CMS has yet to give any guidance on how that process will occur, Dreyer says. “Hopefully that will happen in the rulemaking in November. But that is some of the angst I hear from radiology groups, and that is having questions about what they have to prepare for, and they have to tell their billing companies to prepare for.”