Medical malpractice and AI: Jurors react differently depending on how radiologists utilize the technology
As more organizations begin to implement artificial intelligence into their workflows it is important to not only consider the legal liabilities attached to the technology, but how patients perceive these liabilities as well.
One of the biggest questions relative to the use of AI in medical practices centers on liability—when an algorithm makes a mistake, is it the AI at fault or the provider who is utilizing it? This is especially important in radiology, as there are significantly more FDA-cleared AI applications designed for imaging tasks than there are in any other specialty.
A new analysis published in Nature Health delves deeper into this topic, posing an important question that could alter how medical malpractice cases are managed in the future: How do jurors perceive physician liability when the use of AI is at play in diagnosing and treating patients? According to mock trial data, the answer depends on how the AI is utilized.
“If you're a stakeholder trying to figure out whether to purchase an AI product at a hospital, whether to direct your doctors to follow a certain workflow, or whether to settle a case because an error has already occurred, this kind of information is vital because you can weigh the cost versus the benefits in a far more informed way,” noted co-author Brian Sheppard, professor of law at Seton Hall University School of Law.
A team of legal and medical experts from Penn State College of Medicine, Brown University and Seton Hall University School of Law conducted a mock trial that entailed a series of hypothetical malpractice cases involving the use of AI in patient care. Mock jurors were recruited to provide their perception of liability in cases where radiologists missed a brain bleed on a CT scan, causing irreversible brain damage.
Nearly 300 jurors were randomized into two groups—one that deliberated on a scenario where AI flagged an exam as abnormal and a radiologist reviewed it afterwards, deeming it to be normal; and one that heard the case of a radiologist reviewing and interpreting the CT twice—once before receiving AI feedback and then a second time after AI flagged the case as abnormal.
In both cases, the radiologist’s final interpretation indicated that there was no brain bleed. Despite the finding being missed in both cases, 75% of the jurors perceived the radiologist to be at fault when they reviewed the CT scan only once after AI had flagged the exam as abnormal. When the radiologist reviewed the scan twice, this number fell to 53%.
These findings could have important implications for workflow, as improving efficiency is one of the biggest perceived benefits of integrating AI into clinical practice. It will be important for practices to assess the tradeoffs for how AI will impact not just efficiency, but how radiologists go about their interpretations alongside AI as well, which could affect the perception of liability in legal cases, the team suggested.
“There are all these biases that are incentivizing radiologists not to disagree with AI because the cost of disagreeing with it is too high. If you disagree with AI and you're wrong, this will be used against you,” said co-author Grayson Baird, associate professor of radiology at Brown University and director of the Brown Radiology, Psychology, and Law Lab and the Brown Radiology Human Factors Lab. “The cost is then passed on to the patient who now has to deal with the anxiety and discomfort from follow-up care, imaging or tests. We all pay for it, too, because the cost of healthcare increases.”
Read more here.
In addition to her background in journalism, Hannah also has patient-facing experience in clinical settings, having spent more than 12 years working as a registered rad tech. She began covering the medical imaging industry for Innovate Healthcare in 2021.