Key Findings from Recent Research

• Bruno and a team from Brown University and Seton Hall University School of Law discovered that the integration of AI into clinical workflows affects how physician liability is perceived.
• Their study, published in Nature Health, involved mock jurors evaluating a hypothetical malpractice case where a radiologist failed to detect a brain bleed in a CT scan, despite AI flagging the scan as abnormal.
• Jurors were nearly 50% more likely to side with the plaintiff when the radiologist reviewed the scan only once after AI feedback, compared to when the radiologist reviewed it twice.

Implications for Healthcare Stakeholders

This research highlights the importance of understanding how AI is integrated into clinical workflows. As Brian Sheppard, a law professor at Seton Hall University, noted, this information is crucial for stakeholders making decisions about AI investments, clinical workflows, and malpractice settlements.

Study Methodology

The study involved 282 participants who were randomly assigned to read one of two scenarios regarding a radiologist’s review of a CT scan. The findings indicated:

1. Approximately 75% of jurors believed the radiologist did not meet their duty of care when reviewing the CT scan once.
2. This percentage dropped to 53% when the radiologist reviewed the scan twice.

Challenges and Considerations

The researchers pointed out that biases may discourage radiologists from disagreeing with AI recommendations due to the potential legal repercussions. This could lead to increased costs for patients and the healthcare system as a whole.

The study underscores the evolving nature of liability perceptions in the context of AI in healthcare, emphasizing the need for ongoing research and adaptation of legal frameworks to keep pace with technological advancements.

For further details, you can refer to the original articles from Penn State University and Penn State Health News.