Development and validation of an interpretable machine learning model for predicting moderate-to-severe urinary retention risk after mixed hemorrhoidectomy

Abstract Objective: To develop and validate an interpretable machine learning (ML) model for early prediction of moderate-to-severe postoperative urinary retention (POUR) after mixed hemorrhoidectomy. Methods: Data from 410 patients (training/internal validation) were retrospectively collected (May 2024–March 2025), and data from 261 patients (external validation) were prospectively collected (April–August 2025). Six ML algorithms (GLM, RF, XGBoost, LASSO, SVM, KNN) were compared using AUC, Brier score, Hosmer-Lemeshow test and DCA. SHAP analysis provided model interpretability. Results: Among the six ML models, the LASSO model performed best. This model accurately predicted the risk of moderate-to-severe POUR in patients with mixed hemorrhoidectomy in both the training set (AUC: 0.881, Specificity: 0.979, F1 Score: 0.466) and the external validation set (AUC: 0.739, Specificity: 0.888, F1 Score: 0.333), Brier score=0.150, Hosmer-Lemeshow P <0.001, DCA indicated that the LASSO model provided a higher net clinical benefit. SHAP analysis identified the features contributing to the model predictions, 16 variables were included in the final LASSO model: age, glucose, preoperative fasting, insomnia, operative time, constipation, number of incisions, history of urological diseases, urine_white blood cell (U_WBC), analgesic B (Parecoxib), urine_red blood cell (U_RBC), analgesic C (Dezocine), serum calcium, preemptive analgesia, serum potassium, urine glucose. An online calculator is available at https://mrtools.shinyapps.io/shiny3/. Conclusion: An interpretable LASSO-based ML model was successfully developed and externally validated for predicting moderate-to-severe POUR after mixed hemorrhoidectomy. It enables early identification of high-risk patients and supports individualized perioperative management.

Authors

Institutions

Publication Details

Journal
Research Square
Published
2026-11-17
DOI
https://doi.org/10.21203/rs.3.rs-9891659/v1
Primary Topic
Artificial Intelligence in Healthcare
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Development and validation of an interpretable machine learning model for predicting moderate-to-severe urinary retention risk after mixed hemorrhoidectomy

Long Peng, Minmin Xu, Pingping He, Dong Chen et al.
3 links
Research Square
Artificial Intelligence in Healthcare
preprint

Development and validation of an interpretable machine learning model for predicting moderate-to-severe urinary retention risk after mixed hemorrhoidectomy

Long Peng, Minmin Xu, Pingping He, Dong Chen, Lu Ze
preprint en

Abstract

Abstract Objective: To develop and validate an interpretable machine learning (ML) model for early prediction of moderate-to-severe postoperative urinary retention (POUR) after mixed hemorrhoidectomy. Methods: Data from 410 patients (training/internal validation) were retrospectively collected (May 2024–March 2025), and data from 261 patients (external validation) were prospectively collected (April–August 2025). Six ML algorithms (GLM, RF, XGBoost, LASSO, SVM, KNN) were compared using AUC, Brier score, Hosmer-Lemeshow test and DCA. SHAP analysis provided model interpretability. Results: Among the six ML models, the LASSO model performed best. This model accurately predicted the risk of moderate-to-severe POUR in patients with mixed hemorrhoidectomy in both the training set (AUC: 0.881, Specificity: 0.979, F1 Score: 0.466) and the external validation set (AUC: 0.739, Specificity: 0.888, F1 Score: 0.333), Brier score=0.150, Hosmer-Lemeshow P <0.001, DCA indicated that the LASSO model provided a higher net clinical benefit. SHAP analysis identified the features contributing to the model predictions, 16 variables were included in the final LASSO model: age, glucose, preoperative fasting, insomnia, operative time, constipation, number of incisions, history of urological diseases, urine_white blood cell (U_WBC), analgesic B (Parecoxib), urine_red blood cell (U_RBC), analgesic C (Dezocine), serum calcium, preemptive analgesia, serum potassium, urine glucose. An online calculator is available at https://mrtools.shinyapps.io/shiny3/. Conclusion: An interpretable LASSO-based ML model was successfully developed and externally validated for predicting moderate-to-severe POUR after mixed hemorrhoidectomy. It enables early identification of high-risk patients and supports individualized perioperative management.

Research Square
Nanjing University of Chinese Medicine (CN)
Good health and well-being
Artificial Intelligence in Healthcare
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon