Key result
A data-driven prediction model using preoperative data accurately predicted 30-day, 90-day, and 1-year mortality after colorectal cancer surgery with AUROCs of 0.88, 0.878, and 0.861, respectively.
Why the study?
Existing postoperative mortality risk scores after colorectal cancer resection rely on intra- or postoperative variables or fail to represent contemporary clinical practice.
Population
57,558 patients undergoing colorectal cancer surgery across four Danish databases
Comparison
Preoperative data-driven prediction models for 30-day, 90-day, and 1-year mortality
Design
Observational database-derived prediction model development study
Follow-up
1 year
Authors
Loading...
Models for post-CRC surgery mortality may aid risk stratification; leaves open external validation and clinical utility.
Observational (n=57,558)
Yes
A data-driven machine learning approach using nationwide health databases successfully developed highly discriminative and well-calibrated prediction models for short-term mortality after colorectal cancer surgery.
Bräuner et al. (2023) conducted an observational in Colorectal cancer (n=57,558). Preoperative data-driven prediction model (LASSO logistic regression) was evaluated on 30-day, 90-day, and 1-year all-cause mortality. A data-driven prediction model using preoperative data accurately predicted 30-day, 90-day, and 1-year mortality after colorectal cancer surgery with AUROCs of 0.88, 0.878, and 0.861, respectively.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: