AI Quality Assurance Models Save Lives and Millions in Avoided Med-mal

He will most likely not be seen, and will likely apply for medical compensation

Aortic dissection
Epidural fluid collections
ruptured spleen
superior mesenteric artery occlusion
intracranial hemorrhage
pulmonary embolism

These are outcomes that are most likely not to be seen, can be devastating to patients, are more likely to be prosecuted, and will be the most costly if they do. I know this because I analyzed it in great detail 220 lawsuits against vRad radiologists between 2017-2020. It is this medical data, along with a massive QA database dating back nearly 20 years, that has guided our team to build AI models for these six specific diseases. We look for potential outcomes that could be seriously life-threatening if we miss – and then use artificial intelligence to help us prevent this from happening.

Lifesaving results

Let’s take a look at the dramatic impact of these models on patient outcomes.

As you can see from the table below, pulmonary embolism And the Intracranial hemorrhage (ICH) are the largest number of diseases identified by our Quality Assurance review models – which is understandable given that we read 6 million studies annually, approximately 85% of which are from emergency departments.

I would like to draw your attention to the “Standard Error Rate” column, which can be useful to guide your vigilance and inform the content of your tutorial. We can calculate these numbers with a high degree of confidence based on actual Diagnosed results as well as missing results identified by AI and our traditional QA bypasses.

missedfindings.png

Aortic dissection And the Epidural spinal fluid These are the two models I chose to present at this year’s SIIM Annual Meeting because they are our longest-running QA models that provide the most data. By summing up all the missing results captured through standard quality assurance bypasses, AI detection, and medical conditions, we can normalize the data to determine that 1% of aortic dissections are not seen by a radiologist and (grab your hats) a whopping 49% From pools of epidural cerebrospinal fluid.

Now at this point in my lecture I can feel people asking, “What’s wrong with vRad radiologists that they’re missing so many epidural fluid sets?” I’m quick to point out, though, that our accuracy might be so better More than others, because our radiologists are specialists in emergency radiology and are more attuned to these findings than a generalist who doesn’t see the volume of emergency cases we do. These are attractive numbers that should guide efforts to improve all radiology practices.

With the loss rate and the total study volume, we can then calculate the annual amount of average cost saved based on the actual physical compensation paid over the years. We save $1,145,000 annually for an aortic dissection and another $728,000 for an epidural using artificial intelligence to communicate missing results to the Quality Assurance Committee faster after the report is signed. Empowering the QA team with the latest technology leads to better patient outcomes and far from litigation in the future.

Critical Epidural Lesions: Last December, vRad data scientist and machine learning engineer, Robert Harris, PhD, presented at the annual meeting of the Radiological Society of North America on the results of our AI model for critical epidural lesions – including epidural hematoma, epidural abscess, and phlegmon Epidural, collection of epidural fluids.

Critical epidural lesions can be catastrophic if undetected, and our model currently finds 30 to 40 per year that are not seen by radiologists.

While less common, SMA انسداد blockage And the spleen ruptures It can have serious consequences for a patient who goes home without treatment, which makes these 8 monthly losses worth identifying.

A patient-first approach to AI with benefits for all

As the radiology industry faces a large volume and shortage of radiologists nationwide, health care providers need to use every tool at their disposal to provide patients with the best possible opportunity for early and effective diagnosis and treatment.

AI has proven to be a highly effective partner for our physicians, and a safety net that is an essential (and complementary) component of critical care provided by emergency diagnostic radiologists.

If you are a radiologist or a group interested in learning more about reading with an AI safety net, I recommend speaking with a physician recruitment officer or account managers who can. I called here.

Leave a Comment