As challenging as it is just to keep up with the shifts around imaging informatics being required right now of healthcare and healthcare IT leaders, peering into the future will perforce prove very daunting for many. Yet some industry thought-leaders are daring to peer forward nonetheless. One who is happy to do so is Rasu Shrestha, M.D., vice president for medical information technology at the vast, 20-plus-hospital University of Pittsburgh Medical Center (UPMC) health system.
Indeed, Shrestha sees imaging informatics from a broad perspective that incorporates both the concept of value-based purchasing in healthcare, and the changing roles of practicing radiologists.Not surprisingly, he speaks as one who is himself a radiologist. Thus, when asked where the U.S. healthcare system needs to be five years and more from now when it comes to diagnostic image management, he sees the need for transformative change on that broader level that encompasses changes to radiology practice itself. “I truly believe that everything that’s been done in the last 10 or so years has been around volume-based imaging, and around optimizing our workflow to the T,” Shrestha says. “That was conscious, and useful; the reality of PACS [picture archiving and communications systems] and RIS [radiology information systems] and 3D and reporting was that we needed to make all that as productive as possible, in the last 10 or 15 years. Where it’s going is to a new paradigm of value-based imaging.”
IN WORKFLOW, A CHANGE IN FOCUS FROM IMAGE TO PATIENT
Shrestha goes on to note that, “Up to now, we’ve been very heavily focused [as practicing radiologists] on dealing with a series of images at a time. We’ve improved our efficiencies, but where things are moving is away from an image-centric workflow and towards a patient-centric workflow, towards treating the patient as a whole. That’s where it needs to go, in terms of value-based imaging,” by which he means radiologists, radiology departments, and the IT and other professionals who facilitate such work, “significantly demonstrating the value that [they] bring to the healthcare enterprise,” far beyond speed and volume capacity of imaging and of diagnostic study report production.
Most importantly, Shrestha notes, “Before PACS, we actually used to be part of the care team, when we were working on film. The surgeon would come down and talk with me about Mrs. Smith. Today, that essentially gets boiled down to an HL7 message through CPOE [computerized physician order entry], and we run the risk of being delegated to the darkest corners of the reading room.” At UPMC, he says, “We’re looking at a number of metrics that might contribute to the value score” of radiologists’ work product, among them, measures of “visibility and transparency,” both related to clinical teamwork. Where IT will be essential going forward, he adds, will be in developing useful measures of radiologist work product that go beyond “volume, speed, and cost.”
Fran Turisco, a principal with the Denver-based Aspen Advisors, says she agrees wholeheartedly with Shrestha about the modifications that radiologists will need to make to their role in the new healthcare. “They will have to think about what their value is to the healthcare equation,” says the Boston-based Turisco, “because in a way, technology is providing such images and tools that are really allowing the referring physicians to learn a lot. I think [Dr. Shrestha is] right that they have to get into the value equation, as opposed to just saying, here’s what I saw on this image—less of a scientific review of images, and more of a care delivery view.”
Meanwhile, when it comes to the technology available to help patient care organizations store the ever-growing waves of images themselves, Turisco notes that “I think what technology is doing is that, in terms of the compression algorithms, they’re getting better and better. No one wants to get rid of anything. The whole idea now is to keep everything forever. I think there are a number of PACS and storage vendors who have worked on improving storage in general, and improving the compression algorithms, so that when they’re stored, they take up less and less space, and when they’re ‘unzipped,’ you have true diagnostic quality. The compression algorithms are being worked out, and at this point, they aren’t worrying about bandwidth as much anymore. We’re worrying less about bandwidth and storage.”
BEYOND IMAGES, CDS THAT WILL PROVIDE POINT-OF-CARE SUPPORT
George T. “Buddy” Hickman, executive vice president and CIO at Albany (N.Y.) Medical Center, and the chairman of the board of the Ann Arbor, Mich.-based College of Healthcare Information Management Executives (CHIME), sees clinical decision support moving to a whole new plane, as thinking computers like Watson pull data and image inputs from a variety of sources in enterprise-wide image and data repositories and use those inputs to support physicians’ diagnoses and decision-making at the point of care. “Think about imaging in the form of radiography and related elements like ultrasound, and then add dermatology images, and then add cellular, pathology images, and think about the data associated with those images,” Hickman says. “What’s going to happen is that those kinds of inputs will be fed into these big computers like Watson, and Watson is going to learn what those images are, and will be able to say when presented with an image, there’s 94 percent chance that this is ‘A,’ meaning a particular condition. So I think the clinician will be doing an ‘over-read,’ whether it’s the pathologist, dermatologist, OB/gyn, or radiologist.”
Hickman immediately adds that, “Of course, there’s no replacement all for the human skill of interpretation. But the initial read will change, based on technology. It won’t happen in five years, but that kind of work is already being done on the cellular level” in some developmental situations now. “So I’m quite sure that’s where things are headed.”
Russ Branzell, president and CEO of CHIME, says that the federal government must move forward rapidly on data standards in order to facilitate this kind of work. Still, “On the micro level, some individual patient organizations are already making progress in that area,” Branzell says, speaking of Poudre Valley Health System (now part of the University of Colorado health system), where we was CIO. “We did this in the organization I was in—radiology, cardiology, pathology, OR films, photos, essentially anything that can be saved as a wav file, image file, or DICOM file, we put into the same storage system. So any primary care physician, trauma surgeon, or OR physician, can then see things. What you see there is, you can reduce the care cycle, the time lag involved in caring for the patient—all of that while reducing the cost, because of reductions in excessive testing. And we were sharing images with most of the health systems in Colorado.” Ultimately, he says, it will be extremely important for healthcare leaders to come together across all the medical specialties, not just radiology, to advance this vision of the easily facilitated sharing of images of all types, and data, for a new world of clinical decision support and collaborative, beam-based care delivery.
