By Jenny Lower
Referring physicians have become accustomed to clinical decision support. But what if a similar tool existed for radiologists—and it could fit in your pocket?
A 2010 study published in Radiology found that only 58% of radiologists who responded were complying with Fleischner Society guidelines for small pulmonary nodules, one of the fundamental standards for radiologists. Roy Kwak, 38, a radiologist at Focus Medical Imaging in Los Angeles, is trying to change that. Kwak is the force behind RadsBest, a free, dynamic iOS app that helps radiologists process their findings and recommend appropriate management and follow-up. “There’s kind of a lag between when something is published and when it’s accepted into the zeitgeist among the radiology community,” Kwak said. “My app is trying to short circuit the time it takes for literature to permeate people’s mental space.”
Kwak refers to his iPhone screen as his “fifth monitor.” His tool works like a calculator, he says, consolidating difficult algorithms from published standards into a user-friendly series of questions. In the case of an asymptomatic pancreatic cystic mass, the app translates the detailed algorithm from the ACR white paper on the subject into speedy clickable questions. What size is the mass? What are its imaging characteristics? Has it changed in size since a prior exam? No growth means the mass is benign and requires no further follow-up; a mass greater than 4 cm means the patient should be referred for surgery. If a question won’t materially affect the outcome, it drops out of the sequence, letting the user reach the answer faster.
A dynamic back end allows Kwak to tweak the interface and upload new standards as they become available. The app can be downloaded to a mobile device, so that even radiologists working in environments with poor Wi-Fi can still refer to it. And—likely a popular feature—users download the app directly, without awaiting a stamp of approval from the hospital IT department.
According to Kwak, his app helps radiologists stay up-to-date on the latest research and standards. What’s more, radiologists can pass along the appropriate imaging-related follow-up steps to the referring physician, making their reports more actionable and earning a more central role in patient care. Kwak sees his tool as part of a larger trend like that of Imaging 3.0, the American College of Radiology initiative that urges radiologists to embrace technology and position themselves as expert consultants and instrumental in helping to deliver quality patient care in an evolving healthcare system. “If all the radiologists are using the guidelines, then theoretically their reports and recommendations will be more consistent, and the quality will be better,” Kwak said.
The idea first struck Kwak in 2010 while he was working as an attending radiologist in diagnostic radiology for the New York Harbor Veterans Affairs Hospital in Manhattan. His department chief had recommended a paper by Lincoln Berland about managing incidental findings on CT scans of the abdomen and pelvis. Despite a complementary set of algorithms and flow charts, wading through the details was “kind of a beast,” Kwak said.
Though he lacks a programming background, Kwak spent his off hours exploring basic methods. The first version of RadsBest (short for Radiologist’s Best Friend) went live in the iTunes store in June 2013. To date, it has received 400 downloads—a respectable number, until you consider that there are some 30,000 radiologists in the United States alone. Kwak wants to reach all of them. To that end, he’s working on an updated version of RadsBest that will allow users to log in through social media sites like Doximity, LinkedIn, and Facebook. Version 1.1 will contain some features for purchase, with the proceeds used for supporting development costs for new modules, but the app itself—along with a significant chunk of content and ongoing updates—will remain free.
For more information, visit www.radsbest.com.