by Sean Ruck
, Contributing Editor | May 11, 2018
From the May 2018 issue of HealthCare Business News magazine
The current state of health information exchange generally excludes imaging, and when it does include it, it is flawed, in that it doesn’t integrate with a radiologist’s work flow.
There are three broad ways to exchange health information, according to Nicola. The first is patient-driven, meaning patients have full access to whom they deliver their health information to, including prior imaging. The second is physician-directed, where the referring physician has the option to share images with radiologists. The third is the query-based exchange, which utilizes database queries to find anything done on a patient. The first two are essentially non-starters for radiologists. “The reason why those two types of exchanges don’t work particularly well in radiology is because very few people understand what could prevent the next study in radiology, except for the radiologist,” Nicola said.
He offered an example; “There are times that I might see an abnormal kidney mass on a CT abdomen/pelvis, but an MR lumbar-spine from 10 years ago could prevent a follow-up imaging and I don’t think patients or doctors connect, ‘oh this lumbar MR might help on a CT with kidneys.’ They don’t connect that, but it turns out that occasionally, the MR picks up a little bit of the kidney and if you see the mass from 10 years ago, you know it’s probably nothing to worry about. So really, the radiologist needs to have access to all of the patient’s prior imaging, and they need to be able to acquire it themselves, with the patient’s permission, but it has to be us who are really searching for that prior imaging.”
Ideally, Nicola said, that search would be done automatically with matching paradigms and protocols in place, so it can be incorporated into the radiologist’s workflow, as opposed to having to sit at a separate workstation and actively pull it. “I think artificial intelligence could play a big role facilitating the query of databases, initiating the process of acquiring prior imaging from a broader database, and delivering those images to PACS prior to our interpreting the case. That would be huge leap for the specialty toward practicing high quality cost-effective care, but it doesn’t exist right now. So we’re stuck with the physician-based or patient-based exchanges that are really not helping our workflow or cutting redundant imaging.”
Even with those challenges, Nicola feels that radiology will be a big factor in value-based care, specifically because of the work being done in standardization of how radiology reports are structured, and how recommendations are made. “I think radiology is at the forefront of the standardization movement,” he said. “Because mainly, it’s a technology-driven specialty. People that go into it rely on technology and understand standardization is vital for data mining. Once you’ve done this, using the full-power of machine learning algorithms will be at our fingertips; we’ll really be able to use data to drive care, the foundation of change management. We will eventually have robust data sets allowing us to recommend the correct treatment or cost-effective standardized follow-up recommendations without the large amount of variability inherent in the current practice of the specialty.“
Back to HCB News