Consider this scenario: a call comes into your clinic from a local primary care physician who has found a suspected abdominal aortic aneurysm in a patient, a 65-year-old man, after a routine examination. You ask the physician to send the patient’s contact information to your care team, who books a follow-up examination. Further imaging of the aneurysm reveals that the aorta is just below the 5.5 cm threshold required to intervene, based on current national AAA screening guidelines.
You explain to the patient that you need to keep an eye on their condition, and you hope that they will come back in six months for a follow-up scan. If your 65-year-old is like 35% of abdominal aortic aneurysm patients, they never return, and, like many, they later turn up with a ruptured aneurysm in the emergency room in their local community hospital.
Ruptured abdominal aortic aneurysm (AAA) is a leading cause of death in America. There are an estimated 1.1 million Americans with an AAA, who are often asymptomatic until rupture. Of those patients who experience a rupture, over 10,000 of them – 85 percent, will die. Frustratingly, 43% of rupture AAA deaths occur in patients who are outside of the national AAA screening guidelines: 34% in women and 9% in men who are under the age of 65.
Despite the urgent need to better identify and treat AAA patients, the vascular care coordination needed to address the condition is in a fractured state. Consider another scenario.
A patient enters your emergency department presenting with severe belly or back pain. A CT scan is taken, and the patient is diagnosed with kidney stones. While the radiologist noted some aortic ectasia, the patient was put into the kidney stone workflow and left off of the AAA worklist. The patient is lost to follow-up and presents years later with a rupture.
In addition to these patients who aren’t identified as being at risk or those who fall through the cracks, there are other factors that raise barriers and complications to care. Patients who are non-compliant, the complexity of health insurance, nursing shortages, or reduction in vascular surgery trainees all mean that the path to connecting vascular care teams with the patients that need the right care at the right time is a daunting challenge.
What can be done to help ensure that at-risk patients do not fall through the cracks? This is one area where artificial intelligence and machine learning (AI/ML) can make contributions to vascular care that will have both immediate and long-term impacts in saving lives. In fact, AI/ML-powered interventions can, through early detection and improved communication lead to elective repair and help prevent AAA rupture.
AI-ML-powered innovation can be harnessed to automatically search for the presence of a suspected abdominal aortic aneurysm from any CTA purely based on data and without discrimination. If a suspected AAA is found, the relevant member of the care team – whether it’s a vascular coordinator, surgeon, fellow, pulmonologist, nurse practitioner or nurse navigator, is notified on a real-time, continual basis. This gives the care team the opportunity and ability to capture and enroll patients into follow-up care, counsel them about their condition, and recommend urgently needed care options, if necessary.
With this in mind, several players are working to incorporate AI into their care modules. At Viz.ai, where I serve as an advisor, they recently received the first FDA clearance for an AI-powered solution for the detection and triage of suspected AAA. This algorithm leverages AI to automatically search for the presence of a suspected abdominal aortic aneurysm (AAA) from any computed tomography angiography (CTA) from any scanner in any institution in a hospital network that uses the company’s platform. If a suspected AAA is found, the relevant member of the care team is automatically notified. This gives the care team the ability to capture and enroll patients into follow-up care and/or counsel them indiscriminately.
It’s part of a movement in care innovation that’s much-needed and growing. AAA rupture is a catastrophic medical emergency, and only half of patients with ruptured AAAs reach the hospital alive. For those who do, the survival rate drops by about 1% per minute. Up to 80% do not survive repair. And the story is not the same for men and women.
The United States Preventive Services Task Force (USPSTF) recommends AAA screening for men aged 65 to 75 years with a history of smoking because of the high prevalence of AAA in this population and the demonstrated reduction in AAA-related mortality among those who have been screened. However, at least 33% of ruptured AAA hospitalizations and 41% of aortic aneurysm deaths are among women, and 22% of AAA-related deaths occur in nonsmokers. For these groups, screening has received a grade D (against) or C (neutral recommendation), respectively, by the USPSTF. Thus, current national guidelines for AAA screening fail to target some of the populations in which many AAA ruptures and deaths occur.
This gap between guidance and actual outcomes is yet another reason modalities of care that identify disease outside recommended parameters are so critical. They can do the work of capturing and referring patients automatically, follow the presence of AAA regardless of patient age, sex and location, and ultimately, prevent catastrophic aortic emergencies.
AI powered solutions are becoming increasingly available and should be welcomed by overstretched, under-resourced vascular teams that need the benefits of care coordination. This solution is a positive step forward for everyone: physicians, care teams and hospital systems, and, most importantly, for their patients.