As radiology struggles to find its footing in an emerging healthcare delivery paradigm that emphasizes collaboration and accountability, radiologist Rasu Shrestha, MD, finds himself at the center of the fray at the University of Pittsburgh Medical Center, Penn., where he was recently named chief innovation officer and president, Technology Development Center.
The new titles represent an evolutionary step from his previous role as that institution’s vice president of medical information technology. “My newer role is in trying to leverage innovation to take healthcare to the next level,” he told ImagingBiz. During a recent interview, Shrestha shared his vision for the next iteration of PACS, one that puts patients front and center.
How do you envision PACS data being used to improve patient care?
Shrestha: In the area of PACS, when we talk about where we’ve been and where we need to go towards in the industry, we are moving away from volume to value, and that’s fine, but what does value mean? It’s time to stop talking about it and start finding solutions that allow us to have value-based imaging solutions in place. The specifics around value are very broad, and they can very quickly go very deep, as well. There’s no doubt that there is a definite shift from a volume-based practice of imaging to one that emphasizes value across the care continuum.
We’ve had tremendous innovation in the last 10, 20 years in imaging. More than any other specialty, we’ve lead by innovating: We were the first to embrace the digital form factor. DICOM has been around since 1983.
We’ve had a number of waves of innovation in imaging informatics, the first big wave being the move from film to digital. We first said, “Our need is to get rid of film.” If you were printing film or storing film, there was a very easy ROI on PACS. Then, for five or more years, we spent a lot of time building targeted interfaces and trying to integrate solutions. There’s been a lot of activity around vendor-neutral archives and add-on modules, but it’s still volume-based imaging.
In the first wave, there was a robust business case, we garnered tremendous efficiencies, but in many ways what we did was replicate our analog workflow. PACS, as we know it today, is still very much film-centric. It is still image-centric: We are still treating a series of images at a time and not the patient. That’s my big gripe with where we are now.
What are the key barriers to patient-centric PACS, and how can they be overcome?
Shrestha: Information is still very silo-ed. All that we have through a measly HL-7 interface is basic information such as “reason for exam: cough, or rule out pneumonia.” In this day and age of cloud computing, Facebook and zero downtimes—that just should not be the case.
There is now a lot of information around the context of these patients. The challenge is that it’s buried in other silos. The opportunity at hand is to look at where we’ve been and say, “Look, can we not evolve this image-centric workflow and paradigm and really focus on a patient-centered approach to care?” How can we break down the barriers to this imaging continuum?
How can we bring a fuller patient context to what we are looking at, so it’s not just “reason for exam: headache,” it’s “this patient has a family history of migraine, this patient actually had three or four relevant studies in the past and, oh, by the way, he had a history of a glioblastoma.” We are able to find all of this because we have systems that are talking to each other and also, as a radiologist, I have a fuller awareness of what is going on with my patient.
The other aspect of what we need in PACS going forward is communication. One side effect, and I call it an adverse effect, of PACS, is we stopped talking. Previously we had this rich conversation between a surgeon and myself. He came down with a stack of films and we’d talk about Mrs. Smith who had a family history of “x” and presented with this, and she had this funny gait walking into the clinic and here is what I think and what do you think? Today it’s an HL-7 message that comes across. Mrs Smith is essentially whittled down to a CT head and neck that appears in my worklist with hardly any information around her condition and relevant history.
It needs to be a lot more in terms of leveraging the specifics of the care collaboration between the ordering physician and the radiologist and the technologist and others involved in the care of that patient. That’s what I mean in terms of patient centeredness, and I think there is so much opportunity for us to bring that out in prime time.
How can web and mobile technologies improve access to radiological data as well as interoperability with other data sources?
Shrestha: Interoperability is key. Whereas maybe seven or eight years ago, interoperability was a “nice to have,” today it’s a “must have.” If systems aren’t talking to each other, if we are not able to look across multiple different systems to pull out relevant information about that patient for which we are ordering a study, protocoling or reading an exam, then we’ve defeated the purpose of this entire movement towards digitization in healthcare, we are back to square one. We are practicing film-based radiology except it is in a digital light-box. Interoperability is absolutely key.
Where mobility plays into this is that in the last five years or so, you go from show to show, from RSNA, to HIMSS, to SIIM, and it seems like every vendor is rushing to get an app out, to have a checkmark in the box that says you’ve got an app, you’ve addressed mobility—in the RFP you’ve checked that box—but it defeats the purpose, because mobility really needs to be very purpose-driven, it needs to be purposeful mobility.
One specific example: It is futile for us to have a PACS, even if it is on a zero-footprint client viewer or thin-client viewer that’s essentially a condensed version of the PACS you have on your three-monitor workstation—except it works on an iPad, or an Android device or a Windows device, for that matter.
If you look at the specifics of critical test-results management, you’ve got a specific finding, a pneumothorax that I’ve found as I’m reviewing a study of a patient that came in with chest pain. I’m able to use the mobility framework to call the ordering physician to send the context: “Exactly here is where you need to put the needle in so the patient can start breathing again.” We are able to leverage the mobility platform to essentially get to that ordering physician or surgeon specifically around the use case of critical results management where we close the loop.
Can you suggest other use cases for mobility?
Shrestha: Other specific examples could be around technologists. Technologists are the unspoken heroes of the radiology world. They do a lot of the work that needs to be brought out from a value-based imaging perspective. We talk about how we have to be patient centric in our work, but how many radiologists actually interact with patients in their work?
When you think about the continuum of care, the opportunity for technologists to really hone in on that patient centeredness, the interaction with our patients, leveraging mobile tools, leveraging communication-collaboration platforms, but also building the specifics of what that patient interaction means to improve quality, improve outcome, improve satisfaction from a patient perspective, those all are really key parameters in this new reality of PACS.
What impact do you think the new web versions of HL7 and DICOM will have?
Shrestha: There’s a fair amount of talk about HL-7 FHIR in the industry right now, and I think it’s really important for us to continue to push for standards-based HL-7 FHIR, or even, in DICOM, to continue to look at ways to make DICOM less chatty. We can allow for the metadata and the image data to be separate, so that when we are doing large searches of large databases, we are able to be a lot more efficient in how we are doing these searches, especially in these environments where we have multiple silos of PACS and HIEs, where there are multiple repositories that need to talk to each other.
Standards that are continuing to evolve need to evolve further and faster, and FHIR is multiple steps in the right direction. This entire notion of REST-ful web service is exactly what we need to enhance some of the interoperability needs in the industry. We’ve had a lot of issues around interoperability. I’m a big supporter of FHIR, and a lot of the things we are building at UPMC embrace newer standards.
Is there something more that vendors and radiology services can do to make radiological data, including images and reports, more accessible and patient-friendly?
Shrestha: If we are talking about patient-centric care, then the patient has to be everywhere in the design of the applications we use: That’s us engaging the patient, making sure the patient is informed, and that the patient shares in the decision-making around what we do in imaging.
Patients need access to their reports, they need access to their images, and we need to make sure it’s all a lot more portable. Patients need to be educated on the specifics of why what we are doing is important. The value of the work that we do needs to be communicated to the patient, and I think that is where we can enjoy more of a relationship with the patient. Patient satisfaction scores will go up, and the number of litigations will hopefully start going down.
If we are scheduling a study, we need to be a lot more cognizant of knowing the patient’s preferences and a lot more articulate in looking at prior studies and patient history—and factor all of that into the ordering and decision-support process as well.