Prediction Model for Same-Day Discharge in Robotic-Assisted Radical Laparoscopic Prostatectomy

Area of Science:

• Urology
• Surgical Oncology
• Health Informatics

Background:

• Robotic-assisted radical prostatectomy (RARP) is a common procedure for prostate cancer.
• Predicting hospital stay duration post-RARP is crucial for optimizing patient care and resource allocation.
• Current methods for identifying patients at risk for prolonged hospitalization are limited.

Purpose of the Study:

• To develop and validate a predictive model for prolonged hospital stay after RARP.
• To identify key preoperative factors influencing hospital stay duration.
• To assist in selecting suitable candidates for same-day discharge after RARP.

Main Methods:

• Retrospective analysis of patients undergoing RARP (January 2013 - December 2022).
• Inclusion of preoperative variables: age, BMI, comorbidities, PSA, Gleason score, surgical approach.
• Logistic regression model developed to predict hospital stay > 24 hours; performance assessed via ROC AUC and cross-validation.

Main Results:

• Significant predictors of prolonged stay: BMI, PSA, Gleason score, surgical approach, and comorbidities.
• Transperitoneal approach associated with increased odds of prolonged stay (OR 4.23, p < 0.00001).
• Cross-validated accuracy of the predictive model was 73.7%.

Conclusions:

• The developed nomogram accurately predicts prolonged hospital stay risk post-RARP.
• Informed surgical planning, patient counseling, and postoperative management can be enhanced.
• The model can aid in identifying same-day discharge candidates and optimizing hospital resource use.