International Journal of Surgery

Scan-based scores improved colon surgery risk prediction

Adding scan-based scores to clinical factors improved prediction of death within one year, with a risk-ranking score of 0.80 versus 0.75 for clinical factors alone.

Paper by Hanxue Gu, Yaqian Chen, Jisoo Lee, +4 more authorsChristopher Mantyh · Duke University
An abdominal cross-section showing bowel and muscle sits beside densely packed muscle fibers, fat-separated fibers, and a small group of upright people.

Our read

Biomarkers
Studied:
Cohort study of 1,621 patients undergoing colectomy, 1-year, against clinical factors alone (Multicenter)
What changed:
Adding scan-based scores improved prediction of death within one year: risk-ranking score 0.80 versus 0.75 with clinical factors alone
The catch:
Model development and validation used patients from the same health system

Editor of record: Boris Djordjevic · AgingResearch.News

Follow @agingbrief

Records from 1,621 colectomy patients—people who had surgery to remove all or part of the colon—formed a retrospective cohort study at three hospitals in one health system. Each had a computed tomography (CT) scan within 90 days before surgery. The researchers developed models in 474 patients and tested them in 1,147 later patients.

Adding scan-based scores of frailty, or vulnerability to physical stress, to clinical factors improved prediction of death within one year. A score measuring how well the model ranked risk rose from 0.75 to 0.80. Adding muscle density adjusted for sex to an established surgical risk calculator increased this score from 0.86 to 0.87. The largest benefit occurred among patients the calculator classified as low risk.

Why it matters

Frailty matters for surgical decisions as people age. The authors concluded that scan-based measurements could help guide decisions before surgery.

Caveats

The development and validation groups came from the same health system, leaving the models’ performance in other health systems open. All patients had a suitable preoperative CT scan.

Where the evidence stands

Colectomy

Humans: a single group. Cells, worms or flies and mice: nothing yet.
  • 1 lab in 1 country

Comments

Be the first to comment.

The paper

Improving surgical risk prediction through integrating automated body composition analysis: a retrospective study on colectomy surgery

Hanxue Gu, Yaqian Chen, Jisoo Lee, +4 more authorsChristopher Mantyh

Duke University; Duke University School of Medicine

International Journal of Surgery, Nov 17, 2026, CC BY-NC

doi.org/10.1097/js9.0000000000005737