How Well Can Preoperative Parameters Predict Postoperative Bleeding After Mini-PCNL? A Machine Learning Study
Wie gut können präoperative Parameter Blutungen nach Mini-PCNL vorhersagen? Eine Machine-Learning-Studie
Journal of Endourology · 05.11.2026
DOI: 10.1177/08927790261502115 ↗Auf einen Blick
Präoperative Variablen sagten postoperative Blutungen nach Mini-PCNL mit guter Diskrimination voraus. Random Forest schnitt am besten ab.
Fragestellung
Wie gut lassen sich postoperative Blutungen nach Mini-PCNL mit ausschließlich präoperativen klinischen Parametern vorhersagen?
Methode
Retrospektive Analyse von 1.515 Patient:innen; Entwicklung von Penalized Logistic Regression, Random Forest, XGBoost und LightGBM mit nested cross-validation; Bewertung auf einem Hold-out-Set.
Ergebnisse
Im Hold-out-Set (303 Patient:innen, 52 Ereignisse) erreichte Random Forest die beste Diskrimination: AUC 0,856 (95 %-KI 0,787–0,925), gefolgt von XGBoost 0,838 und LightGBM 0,825; Logistic Regression 0,722. Random Forest war gut kalibriert.
Einschränkungen
Externe, zeitliche und prospektive Validierung sind erforderlich.
Schlussfolgerung der Autor:innen
Ensemble-Machine-Learning-Modelle können postoperative Blutungen nach Mini-PCNL mit guten, gut kalibrierten Vorhersagen auf Basis präoperativer Variablen allein prognostizieren und waren der logistischen Regression überlegen.
Kernaussagen
- Random Forest: AUC 0,856 (95 %-KI 0,787–0,925)
- Logistische Regression unterlag den Ensemble-Modellen
- Externe, zeitliche und prospektive Validierung erforderlich
Deutsche Zusammenfassung: KI-gestützt aus dem Original-Abstract erstellt; alle Zahlen wurden automatisch mit dem Original abgeglichen. Maßgeblich ist der Original-Abstract.
Kurz gefasst
Fachartikel in der Zeitschrift Journal of Endourology, online erschienen am 05.11.2026. Verfasst von Sami Sekkeli und 7 weiteren Autor:innen. Die Arbeit nennt 16 Literaturangaben.
Abstract (Originalsprache)
Objective
To develop and evaluate machine-learning models for predicting postoperative bleeding after mini-percutaneous nephrolithotomy (mini-PCNL) using preoperative clinical parameters only. Materials and Methods: We retrospectively analyzed 1515 patients who underwent mini-PCNL between January 2014 and January 2026. Postoperative bleeding and a postoperative day-1 hemoglobin decrease ≥2 g/dL occurred in 273 patients (18%). All intra- and postoperative variables were excluded to preserve a genuinely preoperative tool, leaving 15 preoperative predictors. Penalized logistic regression, random forest, Extreme Gradient Boosting (XGBoost), and light gradient boosting machine (LightGBM) were developed under a strictly patient-level scheme with nested cross-validation; all preprocessing and categorical-aware oversampling (SMOTENC) were fitted within training folds only. Discrimination (AUC: with 95% confidence intervals [CIs]), calibration (Brier score, calibration slope), and decision-curve analysis were assessed on an untouched hold-out set, with a sensitivity analysis excluding preoperative hematocrit (HCT).
Results
In the patient-isolated hold-out set (303 patients, 52 events), random forest achieved the best discrimination (AUC: 0.856, 95% CI: 0.787–0.925), followed by XGBoost (0.838) and LightGBM (0.825); all three significantly outperformed penalized logistic regression (0.722; DeLong p ≤ 0.013). Random forest was well calibrated (slope 1.099, Brier 0.098) and showed positive net benefit across clinically relevant thresholds. Discrimination remained high after removing preoperative HCT (AUC: 0.828), indicating a multivariable signal. Preoperative HCT, caliceal access site, and stone size were the most influential predictors.
Conclusion
Ensemble machine-learning models can predict postoperative bleeding after mini-PCNL with good, well-calibrated discrimination from preoperative variables alone, outperforming conventional logistic regression. External, temporal, and prospective validation are required before routine clinical use.
Zitieren
APA
Sekkeli, S., Caglar, U., Sekkeli, M. E., Yildiz, O., Ozgor, F., Sarilar, O., Halis, A., & Aksu, U. C. (2026). How Well Can Preoperative Parameters Predict Postoperative Bleeding After Mini-PCNL? A Machine Learning Study. Journal of Endourology. https://doi.org/10.1177/08927790261502115
Vancouver
Sekkeli S, Caglar U, Sekkeli ME, Yildiz O, Ozgor F, Sarilar O, et al. How Well Can Preoperative Parameters Predict Postoperative Bleeding After Mini-PCNL? A Machine Learning Study. Journal of Endourology. 2026. doi:10.1177/08927790261502115
BibTeX
@article{Sekkeli2026,
author = {Sekkeli, Sami and Caglar, Ufuk and Sekkeli, Muhammed Eren and Yildiz, Oguzhan and Ozgor, Faruk and Sarilar, Omer and Halis, Ahmet and Aksu, Ufuk Can},
title = {How Well Can Preoperative Parameters Predict Postoperative Bleeding After Mini-PCNL? A Machine Learning Study},
journal = {Journal of Endourology},
year = {2026},
doi = {10.1177/08927790261502115}
}
EUVIMED