Sealing the Cracks: Re-Envisioning Radiology’s Role in Preventing Loss to Follow-up

With a double-click, my cursor transforms into a spinning blue circle while the next study, a CT scan of the abdomen, buffers and loads. The images display within squares across the monitor, thinly sliced and grouped. I begin to sift through and report on the findings… “Lung bases are clear. Liver and spleen are normal. At the pancreas, infiltrative soft tissue emanates from the gland and encases the adjacent structures.” The appearance is that of an aggressive and incurable cancer.

For many pancreatic cancers, the cause remains elusive and the rapidity of growth precludes screening options. In other words, there is often nothing the patient or doctor could have done differently to change the fatal outcome. Having personally diagnosed hundreds of pancreatic malignancies in my 15 year career, I can no longer remember every unfortunate case. However, this case is different.

The patient had a prior radiologic study interpreted by a colleague of mine just 3 months earlier, a CT angiogram designed to evaluate the major blood vessels in the abdomen. While not the intention of the study, the pancreas tumor was incidentally discovered. At that time, the mass was smaller and spared adjacent structures… i.e. surgery would have been curable. 

With such a lethal malignancy, why did the surgeon wait to operate? Did the patient refuse surgery? I dive into the electronic medical record searching for answers… but find none. No mention of the life threatening cancer in the ordering doctor’s clinic note. No reference to the ticking time bomb in the subsequent vascular surgery consultation. No call to arms when the patient presented a month later to an urgent care. The cancer was discovered months ago by a radiologist and reported in time to cure the patient, but no one had read the report. 

In a blatant contrast to ‘do no harm’, a cascade of errors contributed to the death of a patient. The system failed to protect this patient from falling through the cracks and being lost to follow up. Like slices of Swiss cheese, each stage contains holes and sometimes an error slips through. 

About 25% of clinically significant findings in a radiology report never get acted on. With this in mind, the ACR Practice Parameters for the Communication of Diagnostic Imaging state non-routine communication may be warranted if the interpreting physician reasonably believes the findings are significant and unexpected, may have a reasonable probability of impacting the patient’s health, and may not require immediate attention but, if not acted on, may worsen over time and likely result in an adverse patient outcome.

So, should my colleague have been more proactive and called the results to the ordering provider? In retrospect, of course. But, prospectively the decision is more complicated. A radiologist has a list of patients to care for each day and cannot easily stop to call report without causing a delay in care for the others. Further, the ordering provider may be in the OR or work in a busy clinic, where an unexpected call could be disruptive to their workflow. The reality is radiologists often trust the system and assume ordering providers will review the exam results and act accordingly. 

Rad AI Continuity is a follow up solution designed to solve these real world challenges. Artificial Intelligence models identify and track every type of follow-up recommendation, assessing clinical appropriateness based on national consensus guidelines and health system requirements. Then, the model automates imaging study orders, lab orders and referrals, with reminders and repeated attempts until the loop is closed. Communication is a multi-pronged approach, with text messages, direct mail, and EHR-based messages to ordering providers and patients, as well as a mobile-friendly provider portal for physicians to track pending follow-ups and easily confirm orders.

An in-depth integration with any EHR allows for maximum efficiency. Comprehensive dashboards provide real-time population health insights. Automated patient follow-ups provide unequivocal new value, ensuring cancers caught early are acted upon. The result is improved patient outcomes, reduced liability, and decreased costs to the healthcare system. Most importantly, continuity offers groundbreaking technology with a tangible impact measured in lives saved. 

Join the thousands of radiologists who trust Rad AI


Request a demo