Griffin Health combines AI & navigation in lung nodule follow-up program

AI and navigation work together

Before implementing the platform, Griffin had no practical way to review radiology reports across more than 100,000 imaging studies each year and determine which patients might need additional attention.

Christine Cooper, vice president of diagnostic and cancer services at Griffin Health, emphasized that incidental pulmonary nodules were already being documented in radiology reports. AI simply gave the health system a way to identify and track those patients systematically.

"There was a real void in terms of how we could evaluate patients for incidental pulmonary nodules. We didn't have technology," she said.

Inflo Health's AI platform mines data from radiology reports and generates a daily worklist of patients who have documentation of incidental pulmonary nodules in the EHR.

The worklist is reviewed by a diagnostic navigator, a board-certified radiologic technologist, who also oversees Griffin's low-dose CT lung cancer screening program. The navigator examines each patient's medical record to determine whether the recommended follow-up has already occurred, Cooper said. If the patient received care outside Griffin, the navigator can also review records through Connecticut's health information exchange.

If care is already underway, no additional action may be needed. If it's not, the navigator helps close the gap.

For patients with a primary care physician, she contacts the physician's office to help coordinate next steps, Cooper explained. For those without a primary care physician, she contacts the patient directly to help connect them with ongoing care.

The goal is not to duplicate care that's already underway, but to identify people who may have fallen through the cracks. Cooper said the navigator provides an additional layer of support for primary care providers as well, helping ensure the recommended follow-up is not overlooked.

"There's so much information coming at [physicians] that things get missed. You need people to be behind all of that," Cooper said.

Reaching patients beyond traditional screening

The AI-supported program is also helping Griffin Health reach underserved patients. The health system found that the demographic of patients with incidental pulmonary nodules differs from those enrolled in its low-dose CT lung cancer screening program.

Unlike the screening program, which depends on physician referrals and established eligibility criteria, incidental findings come from patients entering the health system through the emergency department, inpatient units and outpatient imaging. As a result, the AI-supported incidental findings program reaches a more diverse patient population, including patients without regular access to care, Cooper said.

Although emergency physicians may recommend additional evaluation and provide referrals, some patients never establish that next connection. Cooper stated that the navigator helps bridge that gap by contacting patients directly, helping them establish care and, when appropriate, connecting eligible patients to Griffin's ongoing low-dose CT lung cancer screening program.

Provider engagement in the technology program

 Looking back, Cooper said one of Griffin's biggest lessons was the importance of engaging providers early.

"We launched our program without involving them," she said of Griffin's primary care physicians. "After the fact, it was really important for us to explain what we were doing and why we were doing it."

She now recommends identifying physician champions, communicating with providers before launch and establishing clear expectations for how incidental findings will be managed.

"Doing this in a vacuum doesn't work," Cooper said. Success depends not only on selecting a scalable AI platform, but also on building the workflows, communication and provider support needed to integrate it into clinical practice. Once those elements are in place, she said, organizations should not hesitate to expand the program to additional types of incidental findings.

As AI gives health systems new ways to identify patients who need follow-up, Cooper believes the real measure of success is whether those patients ultimately receive the care they need. At Griffin, the technology helped transform incidental findings from information buried in radiology reports into patients actively guided toward recommended follow-up.

Elizabeth Stricker, BSN, RN, comes from a nursing and healthcare leadership background, and covers health technology and leadership trends for B2B audiences.