Worcester, Mass.-based UMass Memorial Health Care has added up to $1 million annually to its margins by changing its radiology follow-up care workflow.
“This is one of those rare clinical operations where better patient care and financial return align — and that makes it worth the investment,” Eric Alper, MD, vice president, chief quality officer and chief clinical informatics officer at the system, told Becker’s.
About five years ago, the system created an actionable findings team focused on improving radiology workflow. Its primary focus was improving follow-ups.
“For years, we’ve known about incidental findings in radiology — cases where something unexpected shows up on a scan that was done for a different reason,” Dr. Alper said. “If those findings aren’t followed up properly, things can fall through the cracks, and the patient may return later with a more serious issue. We felt a duty to inform the right provider and ensure the appropriate follow-up happens. At the same time, follow-up imaging improves safety and supports our business model, contributing to revenue while reducing risk. It’s a win-win: improved quality, better safety, lower liability and financial benefit.”
The team used a nurse navigator approach to manage follow-ups and provide warm handoffs to providers.
“Our team acts as a safety net — making sure follow-ups are completed and patients receive the care they need,” Colleen Bolen, manager of quality at the system, told Becker’s. “At the same time, we’re helping to lighten the load for providers, making sure nothing gets missed. It’s a support system for both patients and clinicians.”
But at the beginning, there were a lot of workflow inefficiencies.
“The number of findings we were tracking far exceeded what our staff could manage day to day, even with several full-time team members,” Stephanie Mayberg, DMSc, PA-C, senior director of system quality at the system, told Becker’s. “So we asked the team to evaluate the process using Lean methodology — looking for waste like rework or duplication.”
One inefficiency: mailing about 300 patient notification letters per week. Folding and stuffing envelopes took up about 50% of one person’s job — a position with high turnover. To reduce this labor, the team began leasing a letter folder and sealer that can handle 500 letters a minute. The machine replaced about 20 hours of manual labor each week. The team also reevaluated how many patients needed letters and shifted some notifications to MyChart and other technology.
This, and a number of other workflow changes, has allowed UMass to improve follow-up care dramatically. Follow-up completion at the system has risen from 30% to 60%. Not only does this improve patient safety and reduce the risk of delayed diagnosis, but follow-up imaging contributes between $663,000 and $1 million annually to the system’s margins. That doesn’t include any downstream revenue from surgeries, procedures or specialty visits. The team also handles higher volumes of patient findings now, going from 1,800 to 2,300 cases per month, all while having a smaller team. When the work started, there were 8.5 FTEs on staff. Now, there are 5.4.
“This wasn’t an FTE reduction initiative,” Dr. Mayberg said. “We’ve reallocated FTEs to other areas where they were more needed. In one or two cases, if someone left, we chose not to refill that role because we’d optimized the process so well. So we’re doing more with less, not by eliminating roles, but by increasing efficiency.”
These efficiencies are supported by a number of technology and AI solutions, including Epic’s large language model which helps extract findings from radiology reports to find follow-up needs.
“Overall, current EHR and radiology systems aren’t designed to manage this kind of work well,” Dr. Alper said. “We’ve been working with Epic and other vendors to improve their systems, but there’s still room for growth, especially in automation. Radiology is an early adopter of AI, particularly in image recognition, like detecting pulmonary embolisms or strokes. But I think there’s growing opportunity to automate the follow-up process: identifying findings and ensuring timely communication and tracking. That’s where AI can continue to help.”
