Related Experiment Video
Updated: Mar 14, 2026
Competing-Risk Nomogram for Predicting Cancer-Specific Survival in Multiple Primary Colorectal Cancer Patients after Surgery
Published on: September 27, 2024
Development of a Nonparametric Predictive Model for Readmission Risk in Elderly Adults After Colon and Rectal Cancer
Heather Yeo1,2, Jialin Mao2, Jonathan S Abelson1
1Department of Surgery, Weill Medical College, New York-Presbyterian Hospital, Cornell University, New York, New York.
Objectives:
Primary objective: to use advanced nonparametric techniques to determine risk factors for readmission after colorectal cancer surgery in elderly adults.
Secondary Objective:
to compare this methodology with traditional parametric methods.
Design:
Using data from the American College of Surgeons National Surgical Quality Improvement Program (NSQIP), nonparametric techniques were used to evaluate the risk of readmission in elderly adults undergoing surgery for colorectal cancer in 2011 and 2012.
Setting:
More than 200 hospitals participating in the NSQIP database.
Participants:
Individuals aged 65 and older who underwent surgery for colorectal cancer in 2011 and 2012 (N = 2,117).
Measurements:
Age-stratified robust nonparametric predictive model (classification and regression tree (CART) analysis) of 30-day readmission for elderly adults undergoing surgery for colorectal cancer.
Results:
Recent chemotherapy was the most important predictor of readmission in participants aged 65 to 74, with 20% of those with recent chemotherapy and 11% of with no recent chemotherapy being readmitted. Participants aged 75 to 84 who had recently undergone chemotherapy had a readmission rate of 23%, whereas those with no chemotherapy had a readmission rate of 9%. Being underweight was the greatest predictor of readmission (30%) in participants aged 85 and older. These methods were found to be more robust than traditional logistic regression.
Conclusion:
Specific age-related preoperative factors help predict readmission in elderly adults undergoing colorectal cancer surgery. Results of the nonparametric CART analysis are better than traditional regression analysis and help physicians to clinically stratify based on age. This model may help identify individuals in whom intervention may be helpful in reducing readmission after surgery.
More Related Videos
07:13Comparison of Predictive Performance of Three Lymph Node Staging Systems in Colorectal Signet Ring Cell Carcinoma Based on Machine Learning Model
Published on: April 18, 2025
03:05Author Spotlight: Advancing Early Detection and Treatment of Gastrointestinal Tumors
Published on: February 16, 2024