By Rebecca Montz, EdD, MBA, CNMT, PET, RT(N)(CT), NMTCB RS
Vol. 24 No. 2 P. 14
Complexities surround the follow-up of incidental radiology findings.
Advances in radiology and technology have intensified the detection of incidental findings referred to as incidentalomas. These findings are common but unassociated with an ordering physician’s initial clinical question surrounding a patient’s health. However, reporting and following through on radiologist recommendations pose both pros and cons to patients, radiologists, care teams, and health care systems.
Elliot Fishman, MD, a professor of radiology and radiological science at John Hopkins Medicine, says, “More and more incidentalomas are being found, but the challenge is for radiologists and the care team to determine the risks and rewards for patients following the discovery.” Fishman says incidental findings require a balanced approach and emphasizes the importance of not over or under testing, as both have consequences.
Nuance Communications reports that despite 10% of radiology reports recommending follow-up imaging, at least one-half are not conducted. Additional radiology literature has reported that follow-up recommendations are included in 8% to 37% of radiology reports, with 65% of follow-up recommendations being “lost to follow-up.” This means that many patients suffer from delayed or missed diagnoses resulting in poorer outcomes.
On the operations side, this lack of follow-up not only results in a loss of medical imaging and treatment revenue but also increases the risk and liability to a health care system. Conversely, the discovery and reporting of incidentalomas can promote unnecessary care and increase patients’ anxiety while also increasing health care costs.
Another complexity involves following up with ordering or primary care physicians about radiologist recommendations. The World Health Organization estimates that 3.6 billion exams are performed each year globally. Increasing use of radiology exams, along with significant advances in imaging equipment and processing techniques, will continue to drive an increase in the discovery of incidentalomas, challenging patients, radiologists, care teams, and health care systems. However, the challenges of managing incidentalomas provide opportunities for the innovation of AI software and collaborative communication among health care providers.
Studies report ranges of incidental findings from 20% to as high as 98%. The proportion of incidental findings that received little or no follow-up has been reported to be as high as 65%. Radiologists are challenged with complex decisions on reporting incidentalomas.
“Radiology is a powerful tool, but it can be overutilized, which increases radiation exposure, patient anxiety, and health care cost,” Fishman says.
An incidental finding could potentially lead to the discovery and prompt treatment of a previously unknown malignancy, or it could help physicians decide that the finding is of no clinical significance. In each case, radiologists strive to minimize unnecessary care, radiation exposure, and patient anxiety.
Another challenge involves the communication involved with ordering or primary care physicians following through on radiologists’ recommendations. Diagnostic imaging reports often recommend patients for follow-up evaluations, such as additional imaging or labs. However, unless radiologists manually search for the results of follow-up testing or evaluation, it is impossible for them to know whether an ordering physician completed the recommendation or the results from the follow-up. This may lead to the patient becoming “lost to follow-up” in the system and not receiving the timely care they need to improve their outcome.
However, this situation creates an opportunity to improve interdisciplinary care, communication, and collaboration on follow-up plans to improve patient outcomes and minimize undue patient anxiety. The challenges of following through on reported incidentalomas and closing communication and clinical care gaps across a health care system have not gone unnoticed. Health care AI companies, such as Illuminate and Nuance, have worked to increase collaboration on radiologists’ follow-up recommendations among physicians.
Cole Erdmann, vice president and general manager of Illuminate, says, “Improved management of patients with incidental finding follow-up recommendations is more than a radiology opportunity. It is an opportunity for the entire health care organization to improve patient care and achieve financial objectives through timely imaging and treatment.”
Although there have been developments made in standardizing the management of incidentalomas, as well as the use of improved technology such as AI and tracking programs to manage incidentalomas, challenges remain. In this regard, Erdmann says, “We see radiology as having a responsibility to ‘follow through’ with referring providers to ensure their recommendations are seen and understood, while referring providers have a responsibility to ‘follow-up’ with the patient and radiology.”
Common incidental findings include but are not limited to aneurysms, lung nodules, adrenal masses, and pancreatic cystic lesions. Illuminate estimates that incidental adrenal masses are common and found in 3% to 7% of adults. Although the most frequent type of incidental adrenal mass is benign, in cases where malignancy is present, up to 27% of incidental adrenal lesions represent adrenal metastases. Pancreatic cystic lesions are reported on 2.2% of CT exams and 19.6% of MRI exams.
Technology and software have aided in combating the complexities surrounding incidentalomas and follow through of patient care. This includes identifying patients with explicit and implicit follow-up recommendations who already exist in the EHR, as well as new patients who have had incidental findings identified.
In recent years, the use of Al in radiology has increased to assist radiologists’ workflow and patient management. AI in radiology utilizes machine learning applications and software to search medical records and discover insights that will help improve patient management and outcomes. Al can simulate radiologists’ cognitive functions, such as evaluating, reporting, problem solving, and understanding, to help support clinical decisions. The use of AI can also help the health care team with deciding the best treatment options for patients, as well as tracking patients who are recommended for follow-up care.
Al can also enhance patient education while reducing errors and improving administrative and clinical processes. Research has shown that the implementation of Al has improved patient data accessibility and collaboration among health care providers, which positively impacts patient management.
Illuminate has partnered with physicians to develop EHR applications that help eliminate gaps in care by identifying and risk-stratifying patients who will benefit most from immediate follow-up imaging and treatment. Illuminate Discovery360 incorporates people, processes, and technology to identify patients “lost to follow-up” with a variety of incidentalomas. Discovery360 can find these patients, whether they already exist in the EHR or are newly identified. This enables nurse navigators to investigate patients’ medical histories and provide primary care physicians with vital information, facilitating communication between subspecialty physicians and patients to ensure appropriate and timely treatment is administered.
For example, when pancreatic cystic lesions are reported in a patient’s EHR, Discovery360 manages and tracks the patient over multiple years to ensure timely treatment can be provided and proper clinical surveillance is performed. Mark Perry, MD, an associate professor of radiology at the University of Kansas Medical Center, says, “In addition to cancer screening programs, the appropriate follow-up of significant unsuspected incidental findings found in radiology exams is important to ensure better patient outcomes. Ensuring a systematic approach for proper follow-up or treatment of these incidental findings would also expand the clinical value of radiology.”
Another software tool from Illuminate is ActKnowledge, which completes the loop on patient follow-up. Illuminate ActKnowledge performs tasks such as managing call reports. Key features of lluminate ActKnowledge include monitoring patients identified through incidental lung nodules or abdominal aortic aneurysms, lung screening programs, inferior vena cava filter-removal programs, evaluation of chronic conditions, and abnormal imaging findings.
The tool shortens the time between diagnosis and treatment, monitors at risk patients, ensures follow-up care, and facilities communication of diagnostic and treatment activities within the EHR and care team. “Health care organizations need sophisticated patient management and surveillance tools that make it easy for a nurse navigator to manage patient follow-up and communication with the broader care team,” Erdmann says. “Simple worklists or spreadsheets will not provide the needed utility.”
Nuance has also developed AI software to assist with incidental findings. Nuance PowerScribe Follow-up Manager helps health care systems uncover findings, improve patient and physician communication, and provide comprehensive follow up recommendation care plan tracking. The automated system provides a collaborative solution for radiology teams to close the loop on patient follow-up recommendations.
PowerScribe Follow-up Manager uses advanced language understanding to make it easier for intelligent identification and automatic extraction of follow-up recommendations. The system can help eliminate false positives associated with “no evidence of” descriptions and reject irrelevant findings. This provides a more efficient way to improve follow-up to recommendations without increasing the administrative burden on radiology teams or interfering with patient care.
Saint Joseph Mercy Health System (SJMHS) in Michigan found that only 36% of patients with incidental lung nodules recommended for follow-up ever returned for further evaluation. To focus on follow-up, SJMHS implemented PowerScribe Follow-up Manager to integrate AI technology with their EHR and clinical workflows. Since the activation of the tool, SJMHS now achieves over 98% closure rates on incidental lung nodules recommended for follow-up.
Future Patient Management
Radiologists play a significant role in deciding how to address the uncertainty of incidentalomas, and health care systems need to ensure completion of follow-up recommendations. This requires collaboration from a broad range of health care professionals, along with effective communication.
Studies have shown that both lack of treatment management and overtreatment can lead to poor patient outcomes. The development of tools such as AI patient identification and follow-up management systems are essential for timely management of follow-up care and follow through treatment of incidental findings. Multiple studies have shown the feasibility of AI tracking systems for radiology follow-up recommendations, resulting in significant improvements in follow-up completion rates, timely treatment, improved patient outcomes, increased revenue from imaging, and reduced medicolegal risk. As health care moves forward, Al will have a significant effect on how radiologists and physicians practice.
“Not all patients have explicit follow-up recommendations,” Erdmann says. “This, and the fact that follow-up information resides in the unstructured text of clinical reports, requires sophisticated AI tools to identify and manage these patients. Typical EMR search tools are inadequate for the job and require significant manual effort.”— Rebecca Montz, EdD, MBA, CNMT, PET, RT(N) (CT), NMTCB RS, has worked at the Mayo Clinic Jacksonville and University of Texas MD Anderson Cancer Center in Houston as a nuclear medicine and PET technologist.