top of page
< Back

Distinguishing Signal from Perception: What Video-Based Assessment Reveals About Surgical Decision-Making and Technical Performance

Journal of the American College of Surgeons, 2026

Divya Kewalramani, Joel Grunhut, Jeffrey Jopling, Mayur Narayan

Read at the journal  |  View on PubMed

In brief

A letter responding to Ballo and colleagues' study, in which surgeons reviewing robotic sleeve gastrectomy videos predicted postoperative complications with only 28% accuracy. We argue that this result mixes up two separate questions: whether intraoperative technical performance relates to complications, and whether surgeons can consciously see it on video. Poor prediction may reflect the limits of human perception and of current scoring tools rather than a lack of signal in the video itself, which matters for surgical quality improvement and for building surgical AI.

Key findings

  • Low accuracy when surgeons review edited videos after the fact does not show that the video lacks predictive information. It may reflect limits of human perception and cognitive processing.

  • The finding that technical skill scores (GEARS) did not differ between complicated and uncomplicated cases likely reflects measurement limits: a small sample (n = 25), a score designed to separate novices from experts rather than capture procedure-specific events, video editing that removed potentially relevant segments, and reliance on GEARS as the decision framework.

  • Surgeons' qualitative observations did track outcomes (staple-line issues with leaks, suturing patterns with stenosis), and prediction followed visibility: about 50% accuracy for stenosis, where cause and effect are visible in real time, versus 22% for leaks, which surface days later.

  • Technical execution should be assessed together with the procedure's sequence of decision points, since the consequences of an action depend on pathway-specific context.

How to cite

Kewalramani D, Grunhut J, Jopling J, Narayan M. Distinguishing Signal from Perception: What Video-Based Assessment Reveals About Surgical Decision-Making and Technical Performance. J Am Coll Surg. 2026;242(5):1478-1480. doi:10.1097/XCS.0000000000001860

All publications

©2022 by Joel Grunhut. Proudly created with Wix.com

bottom of page