Development and Validation of a Machine Learning–Based Model for Predicting Textbook Outcome after Minimally Invasive Pancreaticoduodenectomy

Development and Validation of a Machine Learning–Based Model for Predicting Textbook Outcome after Minimally Invasive Pancreaticoduodenectomy | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Development and Validation of a Machine Learning–Based Model for Predicting Textbook Outcome after Minimally Invasive Pancreaticoduodenectomy Pengcheng Ma, Zhichen Jiang, Yuanyu Wang, Ze Jin, Zhiang Zhang, and 2 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8269976/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract Background: Textbook Outcome (TO) reflects overall surgical quality. With the expanding use of minimally invasive pancreaticoduodenectomy (MIPD), reliable prediction of TO is essential. This study aimed to identify predictors of TO after MIPD and develop a machine learning (ML) model. Methods: We retrospectively analyzed 411 patients undergoing MIPD (2017–2023). The Least Absolute Shrinkage and Selection Operator regression was used for feature selection, and ten ML algorithms were compared. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC). Model interpretability was assessed using SHapley Additive exPlanations (SHAP). Results: Of 411 patients, 263 (63.99%) achieved TO. Eight variables were identified as predictive features. Among the ten algorithms, the Random Forest model demonstrated the best discrimination (AUC = 0.86). Conclusions: The Random Forest model accurately predicted TO after MIPD and may assist in individualized preoperative risk stratification and perioperative management. Minimally invasive pancreaticoduodenectomy Textbook outcome Machine learning Random Forest Surgical quality Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 INTRODUCTION With the continuous advancement of minimally invasive surgical techniques and the refinement of surgical instruments, minimally invasive pancreaticoduodenectomy (MIPD)—including both laparoscopic (LPD) and robot-assisted (RPD) approaches—has been increasingly adopted in clinical practice in recent years[ 1 ]. In several high-volume medical centers, MIPD has gradually replaced Open Pancreaticoduodenectomy (OPD) and has emerged as a major surgical procedure for the treatment of pancreatic head and periampullary tumors as well as certain benign conditions. The quality evaluation of MIPD is important for surgeons to improve surgical quality. However, conventional assessment methods typically rely on single postoperative indicators, such as complication rate, mortality, or readmission rate, which are insufficient to comprehensively capture the overall quality of complex surgical procedures. The concept of textbook outcome (TO)—a comprehensive indicator integrating multiple key postoperative parameters—has been proposed as a more comprehensive and objective measure of surgical quality[ 2 ]. Initially introduced in colorectal surgery in 2013, TO was later refined with a pancreas-specific definition based on international expert consensus established in 2020[ 3 , 4 ]. Although previous studies have preliminarily explored factors associated with achieving TO after MIPD, dedicated prediction models remain scarce[ 5 ]. Machine learning (ML) has demonstrated remarkable capability in analyzing multidimensional clinical data with complex and nonlinear interactions. In recent years, ML has been increasingly applied in medical prediction research and has shown superior predictive performance compared with traditional statistical methods[ 6 , 7 ]. Based on single-center clinical data, the present study aims to identify key predictors for achieving TO among patients undergoing MIPD and to develop a ML–based prediction model. METHODS Study design and patient selection This study was a retrospective analysis, including patients who underwent MIPD at the Department of Gastroenterology & Pancreatic Surgery, Zhejiang Province People’s Hospital between January 2017 and December 2023. The exclusion criteria were: (1) presence of distant metastases; (2) surgical procedures were modified intraoperatively; (3) incomplete clinical data. The detailed process of data selection and the study workflow are illustrated in Fig. 1 . This single-center retrospective study was approved by the Ethics Committee of Zhejiang Provincial People's Hospital (Approval No. QT2025149), and the requirement for individual consent was waived due to the retrospective design. The study adhered to the STROBE reporting guidelines and complied with the relevant requirements of the Declaration of Helsinki. Variables and definitions Demographic, clinical, and pathological data were extracted from patients’ medical records. Baseline characteristics included age, gender, body mass index (BMI), Nutritional Risk Screening 2002 (NRS-2002) score[ 8 ], history of previous abdominal surgeries, neoadjuvant therapy, pancreatic duct diameter, tumor location, and relevant serological and pathological examinations. Operative outcomes included operation time, intraoperative blood loss, conversion to open surgery, and intraoperative red blood cell (RBC) transfusion. Operation time was defined as the duration from skin incision to completion of skin closure. Postoperative complications were assessed based on the Clavien-Dindo (CD) classification system[ 9 ], with the highest grade of complication recorded for each patient; Major complications were defined as CD grade III or higher. Postoperative pancreatic fistula (POPF)[ 10 ], postoperative hemorrhage (PPH)[ 11 ] were defined and classified according to the standards of the International Study Group of Pancreatic Surgery (ISGPS). Bile leak (BL)[ 12 ] was defined and graded according to the criteria of the International Hepato-Pancreato-Biliary Association. Additional outcomes evaluated included reoperation, postoperative hospital stay, unplanned readmission, and in-hospital or 30-day mortality. The TO[ 4 ] was defined as the absence of POPF, BL, and PPH (all ISGPS grade B/C), major complications (CD grade ≥ III), readmission, and in-hospital or 30-day mortality. Postoperative treatment If there is no clear evidence of postoperative POPF, BL, or infection, antibiotics are usually discontinued 2 days after operation. Amylase measurements of the drainage fluid were conducted since the first postoperative day, and the drainage tubes were removed if the volume was less than 50 mL/day for three consecutive days and the amylase level was lower than three times the upper normal serum amylase level and had a normal appearance. The discharge criteria include: no requirement for intravenous fluids, ability to consume solid or semi-solid food, no need for or only oral analgesics required, well-healed incision without infection, ability to get out of bed independently, ability to walk at least 250 meters, normal function of major organs, and blood test results close to normal. All patients receive a follow-up within 90 days after operation. It is recommended to check hematology, biochemistry, and tumor markers every 3 months, and to undergo a CT or MRI scan every 6 months. Statistical analysis All statistical analyses and data visualizations were performed using R software (version 4.3.0). Categorical variables were expressed as frequencies and percentages. Continuous variables were expressed as means and standard deviations (SD) if they followed a normal distribution, and as medians and interquartile ranges (IQR) otherwise. For comparisons between groups, two-tailed unpaired t-tests were applied to normally distributed continuous variables, while the Mann–Whitney U test was used for non-normally distributed data. Differences in categorical variables between groups were assessed using the Pearson Chi-square test or Fisher’s exact test, as appropriate. All p-values were based on two-sided statistical analyses, and p < 0.05 was considered statistically significant. For model development and validation, the dataset was randomly divided into a training set (70%) and a validation set (30%). Initial feature selection was conducted using the Least Absolute Shrinkage and Selection Operator (LASSO) regression, which penalizes non-essential variables (coefficients compressed to zero), resolve multicollinearity and prevent overfitting—advantages particularly suited to high-dimensional datasets. Based on the training set and the feature selected by LASSO, ten ML algorithms were applied to construct prediction models: Logistic Regression, Random Forest, Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM), Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Decision Tree, Multi-Layer Perceptron (MLP), Adaptive Boosting (AdaBoost), and Naive Bayes. Model performance was evaluated using the validation set. The area under the receiver operating characteristic curve (AUC) was used to identify the model with the best predictive performance. In addition, SHapley Additive exPlanations (SHAP) values were utilized to visualize and interpret the contribution and importance of each feature to the achievement of the TO. We developed an interactive application featuring a real-time scoring interface, enabling users to input relevant data and instantly receive the predicted probability of achieving TO. RESULT Clinical characteristics A total of 411 patients who underwent minimally invasive pancreaticoduodenectomy (MIPD) were included in this study. Among them, 263 patients (63.99%) achieved TO, while 148 patients (35.01%) were classified as the non-TO group (Table 1 ). The TO group had a significantly shorter postoperative hospital stay compared with the non-TO group (P < 0.001). Analysis of the six parameters of TO revealed that the absence of major complications was the major obstacle to achieve TO after MIPD compared with other parameters (Supplementary Table S1 ). Table 1 Clinical characteristics Variables Total (n = 411) Non-TO (n = 148) TO(n = 263) P Age, years 65.00 (57.00, 71.00) 65.00 (58.00, 71.25) 64.00 (57.00, 71.00) 0.196 Male 0.010 NO 173 (42.09) 50 (33.78) 123 (46.77) Yes 238 (57.91) 98 (66.22) 140 (53.23) Hypertension 0.825 NO 239 (58.15) 85 (57.43) 154 (58.56) Yes 172 (41.85) 63 (42.57) 109 (41.44) Diabetes 0.508 NO 306 (74.45) 113 (76.35) 193 (73.38) Yes 105 (25.55) 35 (23.65) 70 (26.62) Previous abdominal surgery 0.795 NO 292 (71.05) 104 (70.27) 188 (71.48) Yes 119 (28.95) 44 (29.73) 75 (28.52) Pancreatic duct diameter > 3mm < .001 NO 136 (33.09) 81 (54.73) 55 (20.91) Yes 275 (66.91) 67 (45.27) 208 (79.09) Pathology 0.877 Malignant 326 (79.32) 118 (79.73) 208 (79.09) Benign 85 (20.68) 30 (20.27) 55 (20.91) Location 0.018 Other 163 (39.66) 70 (47.30) 93 (35.36) Pancreas 248 (60.34) 78 (52.70) 170 (64.64) BMI, kg/m2 22.43 (20.60, 24.33) 22.48 (20.68, 24.75) 22.39 (20.56, 24.18) 0.305 NRS-2002 score 2.00 (1.00, 4.00) 3.00 (2.00, 4.00) 2.00 (1.00, 4.00) < .001 Hemoglobin, g/L 123.00 (110.00, 134.00) 125.00 (108.00, 135.00) 123.00 (110.00, 134.00) 0.895 WBC, 10^9/L 5.59 (4.66, 6.78) 5.64 (4.76, 6.77) 5.50 (4.54, 6.78) 0.255 Neutrophil, 10^9/L 3.50 (2.70, 4.50) 3.53 (2.81, 4.61) 3.40 (2.67, 4.38) 0.356 Lymphocyte, 10^9/L 1.40 (1.10, 1.79) 1.40 (1.09, 1.76) 1.40 (1.13, 1.80) 0.246 Platelet, 10^9/L 220.00 (174.50, 268.50) 220.00 (175.50, 275.50) 220.00 (174.50, 264.00) 0.966 Monocyte, 10^9/L 0.40 (0.30, 0.50) 0.40 (0.30, 0.50) 0.40 (0.30, 0.50) 0.486 Albumin, g/L 37.00 (34.10, 40.48) 36.80 (33.40, 40.20) 37.30 (34.45, 40.60) 0.381 Total bilirubin, µmol/L 19.55 (11.30, 72.38) 21.70 (11.75, 96.55) 18.00 (11.00, 64.35) 0.036 Direct bilirubin, µmol/L 4.80 (2.30, 35.80) 6.20 (2.30, 51.75) 4.45 (2.32, 31.85) 0.058 AST, U/L 32.00 (20.25, 76.00) 33.50 (21.00, 71.00) 30.00 (20.00, 76.00) 0.694 ALT, U/L 38.00 (17.00, 99.50) 38.50 (17.00, 91.00) 37.00 (17.00, 108.00) 0.987 Total Cholesterol, mmol/L 4.61 (3.64, 5.43) 4.71 (3.66, 5.27) 4.57 (3.64, 5.52) 0.725 HDL-C, mmol/L 0.97 (0.69, 1.19) 0.90 (0.56, 1.13) 1.00 (0.71, 1.19) 0.026 LDL-C, mmol/L 2.86 (2.17, 3.45) 2.91 (2.18, 3.25) 2.85 (2.17, 3.55) 0.504 Creatinine, µmol/L 66.60 (57.80, 78.70) 70.30 (58.50, 83.00) 65.40 (57.73, 75.45) 0.030 Glucose, mmol/L 5.39 (4.80, 6.50) 5.39 (4.80, 6.23) 5.39 (4.81, 6.61) 0.355 Triglyceride, mmol/L 1.33 (0.95, 1.83) 1.33 (1.01, 1.97) 1.34 (0.93, 1.80) 0.178 Uric Acid, µmol/L 265.00 (199.75, 337.00) 276.00 (208.25, 341.50) 262.00 (197.00, 330.75) 0.213 CA199, U/ml 42.65 (12.20, 164.20) 46.25 (14.38, 163.90) 40.30 (11.30, 163.90) 0.693 CA125, U/ml 12.20 (8.20, 19.60) 12.75 (8.00, 20.95) 12.10 (8.30, 18.20) 0.497 CEA, µg/L 2.80 (1.80, 4.10) 2.80 (1.87, 4.10) 2.80 (1.80, 4.10) 0.688 PT, s 11.40 (10.90, 12.00) 11.50 (10.90, 12.20) 11.40 (10.90, 11.90) 0.172 APTT, s 26.40 (25.10, 28.10) 26.45 (25.20, 28.33) 26.30 (24.95, 28.00) 0.301 Fibrinogen, g/L 3.38 (2.77, 4.08) 3.40 (2.78, 4.19) 3.35 (2.76, 4.04) 0.798 D-dimer, µg/L 470.00 (260.00, 845.00) 470.00 (250.00, 890.00) 470.00 (270.00, 840.00) 0.927 Blood loss, ml 200.00 (100.00, 400.00) 200.00 (100.00, 500.00) 200.00 (100.00, 300.00) < .001 Operation Time, min 385.00 (305.00, 450.00) 405.00 (335.00, 477.50) 370.00 (290.00, 420.00) < .001 Approach 0.002 LPD 203 (49.39) 88 (59.46) 115 (43.73) RPD 208 (50.61) 60 (40.54) 148 (56.27) intraoperative RBC transfusion 0.013 NO 276 (67.15) 88 (59.46) 188 (71.48) Yes 135 (32.85) 60 (40.54) 75 (28.52) Conversion to open surgery 0.003 NO 371 (90.27) 125 (84.46) 246 (93.54) Yes 40 (9.73) 23 (15.54) 17 (6.46) Postoperative hospital stay, days 15.00 (11.00, 21.00) 21.50 (15.00, 28.00) 13.00 (11.00, 16.00) < .001 Abbreviation: TO, textbook outcome; BMI, body mass index; NRS-2002 score, Nutritional Risk Screening 2002; WBC, white blood cell; AST, aspartate aminotransferase; ALT, alanine aminotransferase; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; CA19-9, carbohydrate antigen19-9; CEA, carcinoembryonic antigen; CA125, carbohydrate antigen125; PT, prothrombin time; APTT, activated partial thromboplastin time; RPD, robot-assisted pancreatoduodenectomy; LPD, laparoscopic pancreatoduodenectomy; RBC, red blood cell. The TO group had a lower proportion of male patients (P = 0.010), a higher proportion of lesions located in the pancreas (P = 0.018), and a greater proportion of patients with a pancreatic duct diameter > 3 mm (P < 0.001). The NRS-2002 score was significantly higher in the non-TO group (P < 0.001). Regarding laboratory measures, the non-TO group exhibited higher total bilirubin and creatinine levels (P = 0.036 and P = 0.030), whereas the TO group had significantly higher high-density lipoprotein cholesterol levels (P = 0.026). Intraoperative findings showed that the TO group experienced less intraoperative blood loss (P < 0.001) and shorter operative times (P < 0.001). The distribution of surgical approaches differed significantly between the groups, with a higher proportion of robotic surgeries in the TO group (P = 0.002). Furthermore, the TO group had lower rates of intraoperative RBC transfusion (P = 0.013) and conversion to open surgery (P = 0.003). LASSO regression for feature selection A total of 411 patients were randomly divided into a training set (70%, n = 287) and a validation set (30%, n = 124) for model development and performance assessment (Supplementary Table S2 ). To mitigate the effects of collinearity, feature selection was performed using LASSO regression on the training set. By selecting a lambda (λ) value equal to one standard deviation from the minimum lambda, where the error is within one standard error of the minimum, eight variables were identified as the most predictive features (Fig. 2 ). The variables incorporated into the final model were: pancreatic duct diameter > 3 mm, intraoperative blood loss, NRS-2002 score, direct bilirubin (DBIL), triglyceride (TG), carbohydrate antigen 125 (CA125), male, and BMI. These variables exhibited the strongest association with the outcome variable, while maintaining model simplicity and effectively reducing the risk of overfitting. Develop and evaluate the performance of the ML prediction models Using the eight selected predictive variables, we developed and evaluated ML prediction models based on ten different algorithms. Ten classification algorithms were trained on the training set, including Logistic Regression, Random Forest, XGBoost, LightGBM, SVM, KNN, Decision Tree, MLP, AdaBoost, and Naive Bayes. Model performance was evaluated on the validation set using the AUC. Among all models, the Random Forest model achieved the highest discrimination, with an AUC of 0.86 on the validation set (Fig. 3 ). Additionally, the Brier score of the Random Forest model was 0.15, which was lower than those of several other models. Decision curve analysis (DCA) further demonstrated that the Random Forest model provided favorable net benefits for clinical decision-making across a wide range of threshold probabilities (Fig. 4 ). Therefore, the Random Forest model was identified as the optimal algorithm for predicting TO, outperforming other ML algorithms. Table 2 summarizes the performance metrics of all models in the validation set. Table 2 The performance metrics of all models Model AUC Brier score Accuracy Sensitivity Specificity PPV NPV Random Forest 0.86 0.15 0.80 0.77 0.84 0.90 0.68 XGBoost 0.82 0.18 0.72 0.65 0.84 0.88 0.58 SVM 0.83 0.16 0.76 0.77 0.73 0.84 0.65 KNN 0.78 0.18 0.69 0.67 0.71 0.80 0.55 LightGBM 0.84 0.15 0.77 0.84 0.64 0.80 0.69 MLP 0.83 0.19 0.77 0.86 0.62 0.80 0.72 Decision Tree 0.74 0.21 0.74 0.76 0.71 0.82 0.63 Logistic 0.83 0.16 0.77 0.86 0.62 0.80 0.72 AdaBoost 0.78 0.18 0.73 0.75 0.71 0.82 0.62 Naive Bayes 0.83 0.19 0.75 0.68 0.87 0.90 0.61 Abbreviation: SVM, Support Vector Machine; KNN, K-Nearest Neighbors; LightGBM, Light Gradient Boosting Machine; MLP, Multi-Layer Perceptron; AdaBoost, Adaptive Boosting; AUC, area under the receiver operating characteristic curve; NPV, Negative predictive value; PPV, Positive predictive value Model interpretation The SHAP summary plot illustrates the impact of each feature on the prediction model (Fig. 5 ). The included features were ranked according to their mean absolute SHAP values, from highest to lowest, reflecting their relative importance in the model. From the most to the least influential, the features were: pancreatic duct diameter > 3 mm, intraoperative blood loss, NRS-2002 score, DBIL, TG, CA125, male, and BMI. According to the prediction model, the higher the SHAP value of the feature, the more likely it is to achieve TO. Finally, we developed an interactive interface that dynamically presents each patient’s predicted probability of achieving TO based on input variables. SHAP force plots visualize the main contributing factors and their specific effects, improving the model’s interpretability and clinical relevance (Fig. 6 ). DISCUSSION This study systematically evaluated 40 potential predictors of TO and ultimately identified eight key variables: pancreatic duct diameter > 3 mm, intraoperative blood loss, NRS-2002 score, DBIL, TG, CA125, male, and BMI. Based on these variables, ten ML models were developed to predict the achievement of TO after MIPD. Among them, the Random Forest model demonstrated superior predictive performance, with an AUC of 0.86. The advantage of the Random Forest model may be attributed to its strong ability to capture nonlinear relationships, model complex interactions among variables, and handle multicollinearity effectively. Furthermore, SHAP analysis was employed to interpret the prediction mechanism of the Random Forest model, providing insights into the relative contribution of each feature to the model’s decision-making process. Surgical quality not only directly influences patients’ short-term outcomes but is also closely associated with their long-term outcomes[ 13 ]. A single indicator is insufficient to comprehensively reflect the overall quality of a surgical procedure. The concept of the TO, as a comprehensive indicator, provides a more holistic and objective assessment of surgical quality[ 5 ]. This concept was first proposed in 2013 and applied to colorectal surgery, after which it was gradually extended to other surgical specialties[ 3 , 14 , 15 ]. In 2020, an international expert consensus further defined TO in pancreatic surgery[ 4 ]. With the widespread adoption of LPD and RPD worldwide, several studies have explored the factors associated with achieving TO after MIPD[ 16 – 18 ]. However, no dedicated prediction model for TO after MIPD has yet been established. Therefore, this study aimed to develop a ML-based prediction model for TO following MIPD. The proposed model is intended to provide a scientific basis for preoperative risk assessment, surgical strategy optimization, and perioperative management, thereby improving surgical quality and enhancing patient outcomes. Several key predictive features identified in this study have been verified in previous research. For example, Wu et al. reported that pancreatic duct dilation (diameter > 3 mm) was associated with an increased probability of achieving TO after LPD[ 17 ]. Cai et al. showed that among patients with pancreatic duct dilation (diameter > 3 mm), female patients were more likely than males to achieve TO after LPD[ 16 ]. Lee et al. suggested that a BMI > 25 reduced the probability of achieving TO[ 18 ]. These findings further support the stability and reliability of the model in this study. In addition, this study identified several other important predictive features, including intraoperative blood loss, NRS-2002 score, DBIL, TG, CA125. The Shapley value plot showed that intraoperative blood loss was a key factor affecting the achievement of TO after MIPD, and its increase was closely related to a decrease in TO probability. Previous studies have confirmed that increased intraoperative blood loss is significantly associated with POPF, mortality, major complications, and readmission, all of which directly lead to TO failure[ 19 , 20 ]. It is necessary to perform preoperative nutritional assessment for all patients undergoing PD surgery[ 21 ]. The NRS-2002 score reflects patients’ preoperative nutritional status, and a higher score indicates a greater risk of malnutrition. A meta-analysis by Sun et al. showed that higher NRS-2002 scores were significantly associated with increased postoperative complications and mortality in abdominal surgery[ 22 ], explaining the finding in this study that higher scores reduce the probability of achieving TO. It is worth noting that nutritional risk can be intervened through preoperative nutritional support. Xu et al. further confirmed that for high-risk patients undergoing OPD with NRS-2002 > 5, preoperative nutritional support can significantly reduce the incidence of POPF[ 23 ]. Two metabolism-related laboratory measures—DBIL and TG—also have predictive value for TO achievement. Elevated preoperative DBIL suggests biliary obstruction or impaired bile excretion, which can lead to hepatic dysfunction, systemic inflammatory response syndrome, or even neurotoxicity through accumulation in the central nervous system; these pathological processes may reduce the rate of TO achievement[ 24 , 25 ]. High preoperative TG levels are also associated with a lower TO achievement rate. Although direct evidence linking high TG levels to MIPD perioperative outcomes is still limited, existing studies in metabolism and cardiovascular fields suggest that elevated TG may be a risk marker for postoperative adverse outcomes[ 26 , 27 ]. CA125, a high-molecular-weight glycoprotein synthesized by mesothelial cells, can be elevated in various malignant tumors, acute inflammations, and certain physiological conditions[ 28 , 29 ]. CA125 levels reflect, to some extent, the tumor or inflammatory burden of the body. Recent studies have shown that CA125 can also serve as a potential marker of in-hospital mortality risk in patients with sepsis[ 30 ]. Our study revealed a new association between CA125 and perioperative outcomes. However, it is important to acknowledge several limitations of our study. First, as a retrospective study, it inherently carries selection bias, and all data were obtained from a single medical center. Second, the prediction model has only undergone internal validation and has not yet been tested in external cohorts, which to some extent weakens the reliability of our findings. In the future, we plan to conduct prospective studies and multi-center collaborative research to further validate and optimize the model in order to achieve more accurate results. We will also continue to report our research findings. CONCLUSIONS Our study introduced a new machine learning algorithm to predict the probability of achieving TO after MIPD, which demonstrated excellent performance. This model may assist in preoperative risk assessment, surgical strategy optimization, and perioperative management for patients undergoing MIPD, although external validation is still required before it can be widely applied in clinical practice. Declarations Competing interests The authors have no conflicts of interest to declare. Funding This study was supported by the National Science and Technology Major Project (2025ZD0552312), Zhejiang Provincial Medical and Health Technology Program (No. 2023KY517), and Key Project of social welfare program of Zhejiang Science and Technology Department, “Lingyan”Program (2022C03099). Author Contribution Pc.M. , Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization; Zc.J. , Writing – review & editing, Writing – original draft, Resources, Project administration, Methodology, Funding acquisition, Data curation; Yy.W. , Investigation, Formal analysis, Data curation; Z.J. , Data curation; Za.Z. , Data curation; Yp.M. , Writing – review & editing, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization; Ww. J. , Writing – review & editing, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization. Data Availability Deidentified data and the R code for model building and validation will be made available from the corresponding author upon reasonable request after publication, subject to institutional data-sharing policies. References Zhang XP, Xu S, Zhao ZM, Yu GS, Han B, Chen X, Ma YT, Xu ZZ, Liu Z, Li ES, Guo XF, Gao YX, Zhao GD, Lau WY, Liu J, Liu R (2023) Outcomes of Robotic Versus Laparoscopic Pancreatoduodenectomy Following Learning Curves of Surgeons: A Multicenter Study on 2255 Patients. Ann Surg Zhang XJ, Fei H, Guo CG, Sun CY, Li ZF, Li Z, Chen YT, Che X, Zhao DB (2023) Analysis of textbook outcomes for ampullary carcinoma patients following pancreaticoduodenectomy. World J Gastrointest Surg 15:2259–2271 Kolfschoten NE, Kievit J, Gooiker GA, van Leersum NJ, Snijders HS, Eddes EH, Tollenaar RA, Wouters MW, -van de Marang PJ (2013) Focusing on desired outcomes of care after colon cancer resections; hospital variations in 'textbook outcome'. Eur J Surg Oncol 39:156–163 van Roessel S, Mackay TM, van Dieren S, van der Schelling GP, Nieuwenhuijs VB, Bosscha K, van der Harst E, van Dam RM, Liem MSL, Festen S, Stommel MWJ, Roos D, Wit F, Molenaar IQ, de Meijer VE, Kazemier G, de Hingh I, van Santvoort HC, Bonsing BA, Busch OR, Groot Koerkamp B, Besselink MG (2020) Textbook Outcome: Nationwide Analysis of a Novel Quality Measure in Pancreatic Surgery. Ann Surg 271:155–162 Yuan J, Du C, Wu H, Zhong T, Zhai Q, Peng J, Liu N, Li J (2025) Risk factors of failure to achieve textbook outcome in patients after pancreatoduodenectomy: a systematic review and meta-analysis. Int J Surg 111:3093–3106 Zhang Z, Zhao X, Shang M, Xu Q, Wang X, Zhang J, Wang C, Gu Z (2025) Machine learning model for postpancreaticoduodenectomy haemorrhage prediction: an international multicentre cohort study. BMJ Open 15:e096147 Xu J, Chen T, Fang X, Xia L, Pan X (2024) Prediction model of pressure injury occurrence in diabetic patients during ICU hospitalization–XGBoost machine learning model can be interpreted based on SHAP. Intensive Crit Care Nurs 83:103715 Kondrup J, Rasmussen HH, Hamberg O, Stanga Z (2003) Nutritional risk screening (NRS 2002): a new method based on an analysis of controlled clinical trials. Clin Nutr 22:321–336 Clavien PA, Barkun J, de Oliveira ML, Vauthey JN, Dindo D, Schulick RD, de Santibañes E, Pekolj J, Slankamenac K, Bassi C, Graf R, Vonlanthen R, Padbury R, Cameron JL, Makuuchi M (2009) The Clavien-Dindo classification of surgical complications: five-year experience. Ann Surg 250:187–196 Bassi C, Marchegiani G, Dervenis C, Sarr M, Abu Hilal M, Adham M, Allen P, Andersson R, Asbun HJ, Besselink MG, Conlon K, Del Chiaro M, Falconi M, Fernandez-Cruz L, Fernandez-Del Castillo C, Fingerhut A, Friess H, Gouma DJ, Hackert T, Izbicki J, Lillemoe KD, Neoptolemos JP, Olah A, Schulick R, Shrikhande SV, Takada T, Takaori K, Traverso W, Vollmer CM, Wolfgang CL, Yeo CJ, Salvia R, Buchler M (2017) The 2016 update of the International Study Group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 Years After. Surgery 161:584–591 Wente MN, Veit JA, Bassi C, Dervenis C, Fingerhut A, Gouma DJ, Izbicki JR, Neoptolemos JP, Padbury RT, Sarr MG, Yeo CJ, Büchler MW (2007) Postpancreatectomy hemorrhage (PPH): an International Study Group of Pancreatic Surgery (ISGPS) definition. Surgery 142:20–25 Koch M, Garden OJ, Padbury R, Rahbari NN, Adam R, Capussotti L, Fan ST, Yokoyama Y, Crawford M, Makuuchi M, Christophi C, Banting S, Brooke-Smith M, Usatoff V, Nagino M, Maddern G, Hugh TJ, Vauthey JN, Greig P, Rees M, Nimura Y, Figueras J, DeMatteo RP, Büchler MW, Weitz J (2011) Bile leakage after hepatobiliary and pancreatic surgery: a definition and grading of severity by the International Study Group of Liver Surgery. Surgery 149:680–688 Wang H, Hu X, Yin C, Zhou D, Li Z, Ma Z, Zhang H (2024) Association of textbook outcomes with improved survival in pancreatic ductal adenocarcinoma following pancreaticoduodenectomy: a retrospective study. Transl Gastroenterol Hepatol 9:38 Lim C, Llado L, Salloum C, Ramos E, Lopez-Dominguez J, Cachero A, Fabregat J, Azoulay D (2021) Textbook Outcome Following Liver Transplantation. World J Surg 45:3414–3423 Wu MY, McGregor RJ, Scott J, Smithers BM, Thomas J, Frankel A, Barbour A, Thomson I (2023) Textbook outcomes for oesophagectomy: A valid composite measure assessment tool for surgical performance in a specialist unit. Eur J Surg Oncol 49:106897 Cai H, Lu F, Gao P, Zhang M, Wang X, Li Y, Meng L, Peng B, Cai Y (2024) Risk factors of textbook outcome in laparoscopic pancreatoduodenectomy: results from a prospective high-volume center study. BMC Surg 24:233 Wu Y, Peng B, Liu J, Yin X, Tan Z, Liu R, Hong D, Zhao W, Wu H, Chen R, Li D, Huang H, Miao Y, Liu Y, Liang T, Wang W, Yuan J, Li S, Zhang H, Wang M, Qin R (2023) Textbook outcome as a composite outcome measure in laparoscopic pancreaticoduodenectomy: a multicenter retrospective cohort study. Int J Surg 109:374–382 Lee B, Han HS, Yoon YS, Lee JS (2025) Textbook Outcomes of Totally Robotic Versus Totally Laparoscopic Pancreaticoduodenectomy for Periampullary Neoplasm: A Propensity Score-Matched Cohort Study. J Clin Med 14 Seykora TF, Ecker BL, McMillan MT, Maggino L, Beane JD, Fong ZV, Hollis RH, Jamieson NB, Javed AA, Kowalsky SJ, Kunstman JW, Malleo G, Poruk KE, Soares K, Valero V 3rd, Velu LKP, Watkins AA, Vollmer CM Jr (2019) The Beneficial Effects of Minimizing Blood Loss in Pancreatoduodenectomy. Ann Surg 270:147–157 Casciani F, Trudeau MT, Asbun HJ, Ball CG, Bassi C, Behrman SW, Berger AC, Bloomston MP, Callery MP, Christein JD, Falconi M, Fernandez-Del Castillo C, Dillhoff ME, Dickson EJ, Dixon E, Fisher WE, House MG, Hughes SJ, Kent TS, Kunstman JW, Malleo G, Partelli S, Wolfgang CL, Zureikat AH, Vollmer CM (2021) The effect of high intraoperative blood loss on pancreatic fistula development after pancreatoduodenectomy: An international, multi-institutional propensity score matched analysis. Surgery 170:1195–1204 Aleassa EM, Morris-Stiff G (2019) Regarding: Nutritional support and therapy in pancreatic surgery: A position paper of the International Study Group on Pancreatic Surgery (ISGPS). Surgery 165:1248 Sun Z, Kong XJ, Jing X, Deng RJ, Tian ZB (2015) Nutritional Risk Screening 2002 as a Predictor of Postoperative Outcomes in Patients Undergoing Abdominal Surgery: A Systematic Review and Meta-Analysis of Prospective Cohort Studies. PLoS ONE 10:e0132857 Xu JY, Tian XD, Song JH, Chen J, Yang YM, Wei JM (2021) Preoperative Nutrition Support May Reduce the Prevalence of Postoperative Pancreatic Fistula after Open Pancreaticoduodenectomy in Patients with High Nutritional Risk Determined by NRS2002. Biomed Res Int 2021:6691966 Liu JJ, Sun YM, Xu Y, Mei HW, Guo W, Li ZL (2023) Pathophysiological consequences and treatment strategy of obstructive jaundice. World J Gastrointest Surg 15:1262–1276 Pavlidis ET, Pavlidis TE (2018) Pathophysiological consequences of obstructive jaundice and perioperative management. Hepatobiliary Pancreat Dis Int 17:17–21 Chen C, Wen Q, Ma C, Li X, Huang T, Ke J, Gong C, Hei Z (2022) Hypertriglyceridemia is associated with stroke after non-cardiac, non-neurological surgery in the older patients: A nested case-control study. Front Aging Neurosci 14:935934 Murthy NM, Yoo TT, Sanchez A, Chhitu M, Abramov D, Gatling J, Mamas MA, Parwani P (2025) Preoperative Diagnostic Assessment of Patients with Cardiovascular Risk Factors Undergoing Noncardiac Surgery: A 2025 Update. Methodist Debakey Cardiovasc J 21:87–100 Buamah P (2000) Benign conditions associated with raised serum CA-125 concentration. J Surg Oncol 75:264–265 Sjövall K, Nilsson B, Einhorn N (2002) The significance of serum CA 125 elevation in malignant and nonmalignant diseases. Gynecol Oncol 85:175–178 Gomar S, Tejeda L, Bou R, Romero B, Quesada-Dorador A (2022) Association of carbohydrate 125 antigen with sepsis mortality in critical patients. Med Clin (Barc) 159:124–129 Additional Declarations No competing interests reported. Supplementary Files SupplementaryTableS1.docx SupplementaryTableS2.docx Cite Share Download PDF Status: Posted Version 1 posted You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-8269976","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":556736135,"identity":"80482d7b-645c-42a8-9c10-d97e9335f6c2","order_by":0,"name":"Pengcheng Ma","email":"","orcid":"","institution":"Zhejiang Provincial People's Hospital","correspondingAuthor":false,"prefix":"","firstName":"Pengcheng","middleName":"","lastName":"Ma","suffix":""},{"id":556736136,"identity":"7f29575b-553e-4af2-85f5-9351d8b2fe04","order_by":1,"name":"Zhichen Jiang","email":"","orcid":"","institution":"Zhejiang Provincial People's Hospital","correspondingAuthor":false,"prefix":"","firstName":"Zhichen","middleName":"","lastName":"Jiang","suffix":""},{"id":556736137,"identity":"f39e7d37-e147-4e2d-a5a4-6d60e509371c","order_by":2,"name":"Yuanyu Wang","email":"","orcid":"","institution":"Zhejiang Provincial People's Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yuanyu","middleName":"","lastName":"Wang","suffix":""},{"id":556736138,"identity":"bd3f0113-bca0-44e4-8c9d-d3caf64ecd1e","order_by":3,"name":"Ze Jin","email":"","orcid":"","institution":"Zhejiang Provincial People's Hospital","correspondingAuthor":false,"prefix":"","firstName":"Ze","middleName":"","lastName":"Jin","suffix":""},{"id":556736139,"identity":"e0b0afd2-05b9-4238-9b56-b490fa053c55","order_by":4,"name":"Zhiang Zhang","email":"","orcid":"","institution":"Zhejiang Provincial People's Hospital","correspondingAuthor":false,"prefix":"","firstName":"Zhiang","middleName":"","lastName":"Zhang","suffix":""},{"id":556736140,"identity":"1e6f02af-5a49-4c54-a2af-8ff00d3fd5d1","order_by":5,"name":"Yiping Mou","email":"","orcid":"","institution":"Zhejiang Provincial People's Hospital","correspondingAuthor":false,"prefix":"","firstName":"Yiping","middleName":"","lastName":"Mou","suffix":""},{"id":556736141,"identity":"3a0de89e-301f-4bbc-b59c-7a08c090129c","order_by":6,"name":"Weiwei Jin","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAArklEQVRIiWNgGAWjYBACxmYQWcHADKIkSNByhhQtEH1tEJo4LcztvMekeefVsRscYD54m4fBLo8Ih/GlSfNuY2M2OMCWbM3DkFxMhBYes9u523iAWnjMpHkYDiQ2EKdljgRQC/83UrQ0GIBsYSNai/nvP8cSmCUPsxlbzjFIJqzFsP+MseGMmrpkvuPND2+8qbAjQgtURTIkMg0IqQcCeShtR4TaUTAKRsEoGKkAABx7MaLHnExDAAAAAElFTkSuQmCC","orcid":"","institution":"Zhejiang Provincial People's Hospital","correspondingAuthor":true,"prefix":"","firstName":"Weiwei","middleName":"","lastName":"Jin","suffix":""}],"badges":[],"createdAt":"2025-12-03 11:53:48","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8269976/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8269976/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":97882612,"identity":"e5b772ac-64f0-4e3c-90a8-b1e67d01e9ce","added_by":"auto","created_at":"2025-12-10 12:44:31","extension":"docx","order_by":0,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":66054,"visible":true,"origin":"","legend":"","description":"","filename":"article.docx","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/da5d0ff48d9245a8cd6dc906.docx"},{"id":97882608,"identity":"d5f9ed50-8eed-4a28-943d-6602d7034d20","added_by":"auto","created_at":"2025-12-10 12:44:31","extension":"tiff","order_by":1,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":271809,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.tiff","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/933229fad73477a2c3d5ce0c.tiff"},{"id":97882609,"identity":"3f7db157-4de3-4388-8b5c-07ccfcde80a1","added_by":"auto","created_at":"2025-12-10 12:44:31","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":30418,"visible":true,"origin":"","legend":"","description":"","filename":"Table1.docx","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/98f5c0f9e1a2db636fb8b7cf.docx"},{"id":97882614,"identity":"ba15c681-4870-41a9-89ff-05db754fdc94","added_by":"auto","created_at":"2025-12-10 12:44:32","extension":"tiff","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8623558,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2.tiff","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/d3ae4a04a47729eb60559098.tiff"},{"id":97882613,"identity":"e18e06bb-09af-4dc2-a8f4-6660611eaf95","added_by":"auto","created_at":"2025-12-10 12:44:31","extension":"docx","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":16280,"visible":true,"origin":"","legend":"","description":"","filename":"Table2.docx","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/8cac929487c732b25d672bf2.docx"},{"id":97882619,"identity":"4a390db8-ff57-40ba-ab7f-54c247148e19","added_by":"auto","created_at":"2025-12-10 12:44:32","extension":"tiff","order_by":5,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6211882,"visible":true,"origin":"","legend":"","description":"","filename":"Figure3.tiff","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/0be21d90221803a026b732a4.tiff"},{"id":97882627,"identity":"59dacf48-ee1c-4e43-b830-2a67aee44177","added_by":"auto","created_at":"2025-12-10 12:44:32","extension":"tiff","order_by":6,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6211882,"visible":true,"origin":"","legend":"","description":"","filename":"Figure4.tiff","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/2605cdf3f81a7943580911b5.tiff"},{"id":97900210,"identity":"97312d42-19a0-4610-80b1-922ea3699e38","added_by":"auto","created_at":"2025-12-10 15:45:18","extension":"tiff","order_by":7,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8623558,"visible":true,"origin":"","legend":"","description":"","filename":"Figure5.tiff","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/1121cf091038b8b5fe490edd.tiff"},{"id":97882624,"identity":"6e95170e-ba8b-4800-86f0-1db245c1c5e0","added_by":"auto","created_at":"2025-12-10 12:44:32","extension":"tiff","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":317723,"visible":true,"origin":"","legend":"","description":"","filename":"Figure6.tiff","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/5c5ad0f0a8422d8940b22103.tiff"},{"id":97882623,"identity":"0c3ba38c-da4d-49be-aada-a2fe6356f0a8","added_by":"auto","created_at":"2025-12-10 12:44:32","extension":"json","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8365,"visible":true,"origin":"","legend":"","description":"","filename":"33b1bc9983ac4ec5ba9b886add10513c.json","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/cca192131c008681af35ace6.json"},{"id":97900104,"identity":"127b3404-8cff-486d-994e-69c24b423012","added_by":"auto","created_at":"2025-12-10 15:45:13","extension":"docx","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":15386,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS1.docx","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/2ebd6d00d7abbd131b8ee43a.docx"},{"id":97882622,"identity":"7bc2201f-ef1b-40c2-9145-a4c88c3f4b47","added_by":"auto","created_at":"2025-12-10 12:44:32","extension":"docx","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":30502,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS2.docx","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/0226d84ab1c4443f51623e02.docx"},{"id":97899938,"identity":"daa691be-6098-458c-b7bc-084cf208207b","added_by":"auto","created_at":"2025-12-10 15:45:07","extension":"xml","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":132603,"visible":true,"origin":"","legend":"","description":"","filename":"33b1bc9983ac4ec5ba9b886add10513c1enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/74f717eea7a784c154608330.xml"},{"id":97900805,"identity":"7b8c838d-6d8f-4895-92ab-907436811f93","added_by":"auto","created_at":"2025-12-10 15:45:55","extension":"tiff","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":271809,"visible":true,"origin":"","legend":"","description":"","filename":"Figure1.tiff","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/e99f5678c5b473f042a51ea4.tiff"},{"id":97900861,"identity":"29e8020e-1622-43ca-b92c-a8039d36c525","added_by":"auto","created_at":"2025-12-10 15:46:00","extension":"tiff","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8623558,"visible":true,"origin":"","legend":"","description":"","filename":"Figure2.tiff","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/27731b25d59d3ae017f1dd00.tiff"},{"id":97900349,"identity":"04c46a0c-fe01-413b-97d1-30af67ff5aeb","added_by":"auto","created_at":"2025-12-10 15:45:23","extension":"tiff","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6211882,"visible":true,"origin":"","legend":"","description":"","filename":"Figure3.tiff","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/a10cbe8d1dbbe77a61fce664.tiff"},{"id":97899062,"identity":"94838a51-aa5e-446c-a4a7-8eb045319d50","added_by":"auto","created_at":"2025-12-10 15:41:00","extension":"tiff","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":6211882,"visible":true,"origin":"","legend":"","description":"","filename":"Figure4.tiff","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/ac2dd149ddc48e02a8bd291a.tiff"},{"id":97882638,"identity":"ddde7f99-2e78-4ad9-9ab4-dbba7a28b630","added_by":"auto","created_at":"2025-12-10 12:44:32","extension":"tiff","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":8623558,"visible":true,"origin":"","legend":"","description":"","filename":"Figure5.tiff","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/c90ceb7cb48509c68f58e089.tiff"},{"id":97882639,"identity":"869cb828-1697-4c76-8e8d-c81a542ed27a","added_by":"auto","created_at":"2025-12-10 12:44:32","extension":"tiff","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":317723,"visible":true,"origin":"","legend":"","description":"","filename":"Figure6.tiff","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/5a9c6f87d823636672457228.tiff"},{"id":97882634,"identity":"33f842d4-a550-42f1-9566-cd7afcbbf159","added_by":"auto","created_at":"2025-12-10 12:44:32","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":53934,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure1.png","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/e22aed5f53749024ff45a823.png"},{"id":97899198,"identity":"fc00c3a9-b411-4218-96f5-c39bec4cf299","added_by":"auto","created_at":"2025-12-10 15:42:11","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":35078,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure2.png","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/4857a7140f0b345546b763b5.png"},{"id":97900487,"identity":"6553d7e1-7edf-4703-ae9a-4f5012c5fcbf","added_by":"auto","created_at":"2025-12-10 15:45:34","extension":"png","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":60117,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure3.png","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/8e970ba6b1ea8465a7e75f35.png"},{"id":97899931,"identity":"71e5d3da-4d77-4901-a43a-176d01294eb0","added_by":"auto","created_at":"2025-12-10 15:45:06","extension":"png","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":55026,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure4.png","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/baf5f7c7c12408b20871b527.png"},{"id":97882631,"identity":"5bfb945b-25c0-4661-b1bb-3fca13273e75","added_by":"auto","created_at":"2025-12-10 12:44:32","extension":"png","order_by":23,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":42340,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure5.png","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/a079fbbfdc5993e32ac502ef.png"},{"id":97882637,"identity":"5a1a5795-ad84-440e-9c17-857c74c65eed","added_by":"auto","created_at":"2025-12-10 12:44:32","extension":"png","order_by":24,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":77441,"visible":true,"origin":"","legend":"","description":"","filename":"OnlineFigure6.png","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/1ed6661260201a518321b6c6.png"},{"id":97882641,"identity":"0f9d7a54-e78e-4a1a-8ea4-957c424439cf","added_by":"auto","created_at":"2025-12-10 12:44:32","extension":"xml","order_by":25,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":129726,"visible":true,"origin":"","legend":"","description":"","filename":"33b1bc9983ac4ec5ba9b886add10513c1structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/16d5dfcf05b10920e218ccba.xml"},{"id":97882640,"identity":"1d5080e5-f7ec-44e6-b10c-f6c2d2e25910","added_by":"auto","created_at":"2025-12-10 12:44:32","extension":"html","order_by":26,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":137714,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/8040c0c04aee4b0d07ec14a4.html"},{"id":97882606,"identity":"97554d3e-bac0-4281-80a5-bd6d0a48fc07","added_by":"auto","created_at":"2025-12-10 12:44:31","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":1500262,"visible":true,"origin":"","legend":"\u003cp\u003eFlowchart of the study\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/6c2f63c15fcb70a67180422b.png"},{"id":97882607,"identity":"f06b5176-2cee-4c87-b6a4-02d2e673ae3f","added_by":"auto","created_at":"2025-12-10 12:44:31","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":483117,"visible":true,"origin":"","legend":"\u003cp\u003eLASSO regression for feature selection\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/ae0c769f6f35fda7752b123e.png"},{"id":97882611,"identity":"1907f3ae-d173-48df-be44-f027f565a3e2","added_by":"auto","created_at":"2025-12-10 12:44:31","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":1141197,"visible":true,"origin":"","legend":"\u003cp\u003eROC curves for the ten prediction models\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/b1ff223930876799ff1bab20.png"},{"id":97899693,"identity":"7b171b50-c91d-4be8-beaf-41c43cc33e68","added_by":"auto","created_at":"2025-12-10 15:44:50","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":1217382,"visible":true,"origin":"","legend":"\u003cp\u003eDCA curves for the ten prediction models\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/5eb9b733174dbd94d6334859.png"},{"id":97899448,"identity":"d560dc29-2c1c-424c-bda5-3b094c1b9f35","added_by":"auto","created_at":"2025-12-10 15:44:32","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":824012,"visible":true,"origin":"","legend":"\u003cp\u003eSHAP summary plot\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/384dd0730dad882821da26c9.png"},{"id":97882615,"identity":"6d2dce21-3ac4-47d5-a490-9f207142864f","added_by":"auto","created_at":"2025-12-10 12:44:32","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":1288430,"visible":true,"origin":"","legend":"\u003cp\u003eAn application predicting the achievement of TO\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/ebc730b77fab12e65cd310d4.png"},{"id":99789714,"identity":"166d0b3f-3a49-4113-b2e0-ab1c957f4b07","added_by":"auto","created_at":"2026-01-08 12:50:24","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":7543447,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/7b055df6-4a41-40d4-820b-d00426b7bc05.pdf"},{"id":97899430,"identity":"f2664f2e-6bb9-42c0-999c-bcb32ab72c53","added_by":"auto","created_at":"2025-12-10 15:44:28","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":15386,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS1.docx","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/fe94959ba9dfffd81659d86d.docx"},{"id":97899036,"identity":"15f6c7cc-2072-43c3-a6b4-4e997500dcab","added_by":"auto","created_at":"2025-12-10 15:40:52","extension":"docx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":30502,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS2.docx","url":"https://assets-eu.researchsquare.com/files/rs-8269976/v1/af951ed466deccc8b3938069.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Development and Validation of a Machine Learning–Based Model for Predicting Textbook Outcome after Minimally Invasive Pancreaticoduodenectomy","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eWith the continuous advancement of minimally invasive surgical techniques and the refinement of surgical instruments, minimally invasive pancreaticoduodenectomy (MIPD)\u0026mdash;including both laparoscopic (LPD) and robot-assisted (RPD) approaches\u0026mdash;has been increasingly adopted in clinical practice in recent years[\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. In several high-volume medical centers, MIPD has gradually replaced Open Pancreaticoduodenectomy (OPD) and has emerged as a major surgical procedure for the treatment of pancreatic head and periampullary tumors as well as certain benign conditions.\u003c/p\u003e\u003cp\u003eThe quality evaluation of MIPD is important for surgeons to improve surgical quality. However, conventional assessment methods typically rely on single postoperative indicators, such as complication rate, mortality, or readmission rate, which are insufficient to comprehensively capture the overall quality of complex surgical procedures. The concept of textbook outcome (TO)\u0026mdash;a comprehensive indicator integrating multiple key postoperative parameters\u0026mdash;has been proposed as a more comprehensive and objective measure of surgical quality[\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Initially introduced in colorectal surgery in 2013, TO was later refined with a pancreas-specific definition based on international expert consensus established in 2020[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Although previous studies have preliminarily explored factors associated with achieving TO after MIPD, dedicated prediction models remain scarce[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. Machine learning (ML) has demonstrated remarkable capability in analyzing multidimensional clinical data with complex and nonlinear interactions. In recent years, ML has been increasingly applied in medical prediction research and has shown superior predictive performance compared with traditional statistical methods[\u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e6\u003c/span\u003e, \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. Based on single-center clinical data, the present study aims to identify key predictors for achieving TO among patients undergoing MIPD and to develop a ML\u0026ndash;based prediction model.\u003c/p\u003e"},{"header":"METHODS","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003eStudy design and patient selection\u003c/h2\u003e\u003cp\u003eThis study was a retrospective analysis, including patients who underwent MIPD at the Department of Gastroenterology \u0026amp; Pancreatic Surgery, Zhejiang Province People\u0026rsquo;s Hospital between January 2017 and December 2023. The exclusion criteria were: (1) presence of distant metastases; (2) surgical procedures were modified intraoperatively; (3) incomplete clinical data. The detailed process of data selection and the study workflow are illustrated in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e. This single-center retrospective study was approved by the Ethics Committee of Zhejiang Provincial People's Hospital (Approval No. QT2025149), and the requirement for individual consent was waived due to the retrospective design. The study adhered to the STROBE reporting guidelines and complied with the relevant requirements of the Declaration of Helsinki.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eVariables and definitions\u003c/h3\u003e\n\u003cp\u003eDemographic, clinical, and pathological data were extracted from patients\u0026rsquo; medical records. Baseline characteristics included age, gender, body mass index (BMI), Nutritional Risk Screening 2002 (NRS-2002) score[\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e], history of previous abdominal surgeries, neoadjuvant therapy, pancreatic duct diameter, tumor location, and relevant serological and pathological examinations. Operative outcomes included operation time, intraoperative blood loss, conversion to open surgery, and intraoperative red blood cell (RBC) transfusion. Operation time was defined as the duration from skin incision to completion of skin closure. Postoperative complications were assessed based on the Clavien-Dindo (CD) classification system[\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e], with the highest grade of complication recorded for each patient; Major complications were defined as CD grade III or higher. Postoperative pancreatic fistula (POPF)[\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e], postoperative hemorrhage (PPH)[\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e] were defined and classified according to the standards of the International Study Group of Pancreatic Surgery (ISGPS). Bile leak (BL)[\u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e] was defined and graded according to the criteria of the International Hepato-Pancreato-Biliary Association. Additional outcomes evaluated included reoperation, postoperative hospital stay, unplanned readmission, and in-hospital or 30-day mortality. The TO[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e] was defined as the absence of POPF, BL, and PPH (all ISGPS grade B/C), major complications (CD grade\u0026thinsp;\u0026ge;\u0026thinsp;III), readmission, and in-hospital or 30-day mortality.\u003c/p\u003e\n\u003ch3\u003ePostoperative treatment\u003c/h3\u003e\n\u003cp\u003eIf there is no clear evidence of postoperative POPF, BL, or infection, antibiotics are usually discontinued 2 days after operation. Amylase measurements of the drainage fluid were conducted since the first postoperative day, and the drainage tubes were removed if the volume was less than 50 mL/day for three consecutive days and the amylase level was lower than three times the upper normal serum amylase level and had a normal appearance. The discharge criteria include: no requirement for intravenous fluids, ability to consume solid or semi-solid food, no need for or only oral analgesics required, well-healed incision without infection, ability to get out of bed independently, ability to walk at least 250 meters, normal function of major organs, and blood test results close to normal. All patients receive a follow-up within 90 days after operation. It is recommended to check hematology, biochemistry, and tumor markers every 3 months, and to undergo a CT or MRI scan every 6 months.\u003c/p\u003e\u003cdiv id=\"Sec6\" class=\"Section2\"\u003e\u003ch2\u003eStatistical analysis\u003c/h2\u003e\u003cp\u003eAll statistical analyses and data visualizations were performed using R software (version 4.3.0). Categorical variables were expressed as frequencies and percentages. Continuous variables were expressed as means and standard deviations (SD) if they followed a normal distribution, and as medians and interquartile ranges (IQR) otherwise. For comparisons between groups, two-tailed unpaired t-tests were applied to normally distributed continuous variables, while the Mann\u0026ndash;Whitney U test was used for non-normally distributed data. Differences in categorical variables between groups were assessed using the Pearson Chi-square test or Fisher\u0026rsquo;s exact test, as appropriate. All p-values were based on two-sided statistical analyses, and p\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was considered statistically significant.\u003c/p\u003e\u003cp\u003eFor model development and validation, the dataset was randomly divided into a training set (70%) and a validation set (30%). Initial feature selection was conducted using the Least Absolute Shrinkage and Selection Operator (LASSO) regression, which penalizes non-essential variables (coefficients compressed to zero), resolve multicollinearity and prevent overfitting\u0026mdash;advantages particularly suited to high-dimensional datasets. Based on the training set and the feature selected by LASSO, ten ML algorithms were applied to construct prediction models: Logistic Regression, Random Forest, Extreme Gradient Boosting (XGBoost), Light Gradient Boosting Machine (LightGBM), Support Vector Machine (SVM), K-Nearest Neighbors (KNN), Decision Tree, Multi-Layer Perceptron (MLP), Adaptive Boosting (AdaBoost), and Naive Bayes. Model performance was evaluated using the validation set. The area under the receiver operating characteristic curve (AUC) was used to identify the model with the best predictive performance. In addition, SHapley Additive exPlanations (SHAP) values were utilized to visualize and interpret the contribution and importance of each feature to the achievement of the TO. We developed an interactive application featuring a real-time scoring interface, enabling users to input relevant data and instantly receive the predicted probability of achieving TO.\u003c/p\u003e\u003c/div\u003e"},{"header":"RESULT","content":"\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003eClinical characteristics\u003c/h2\u003e\u003cp\u003eA total of 411 patients who underwent minimally invasive pancreaticoduodenectomy (MIPD) were included in this study. Among them, 263 patients (63.99%) achieved TO, while 148 patients (35.01%) were classified as the non-TO group (Table\u0026nbsp;\u003cspan refid=\"Tab1\" class=\"InternalRef\"\u003e1\u003c/span\u003e). The TO group had a significantly shorter postoperative hospital stay compared with the non-TO group (P\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Analysis of the six parameters of TO revealed that the absence of major complications was the major obstacle to achieve TO after MIPD compared with other parameters (Supplementary Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab1\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eClinical characteristics\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"5\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eVariables\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eTotal (n\u0026thinsp;=\u0026thinsp;411)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eNon-TO (n\u0026thinsp;=\u0026thinsp;148)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eTO(n\u0026thinsp;=\u0026thinsp;263)\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003e\u003cem\u003eP\u003c/em\u003e\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAge, years\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e65.00 (57.00, 71.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e65.00 (58.00, 71.25)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e64.00 (57.00, 71.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.196\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMale\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.010\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e173 (42.09)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e50 (33.78)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e123 (46.77)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e238 (57.91)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e98 (66.22)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e140 (53.23)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHypertension\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.825\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e239 (58.15)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e85 (57.43)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e154 (58.56)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e172 (41.85)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e63 (42.57)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e109 (41.44)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDiabetes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.508\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e306 (74.45)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e113 (76.35)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e193 (73.38)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e105 (25.55)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e35 (23.65)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e70 (26.62)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePrevious abdominal surgery\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.795\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e292 (71.05)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e104 (70.27)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e188 (71.48)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e119 (28.95)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e44 (29.73)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e75 (28.52)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePancreatic duct diameter\u0026thinsp;\u0026gt;\u0026thinsp;3mm\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e136 (33.09)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e81 (54.73)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e55 (20.91)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e275 (66.91)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e67 (45.27)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e208 (79.09)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePathology\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.877\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMalignant\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e326 (79.32)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e118 (79.73)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e208 (79.09)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBenign\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e85 (20.68)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e30 (20.27)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e55 (20.91)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLocation\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.018\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOther\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e163 (39.66)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e70 (47.30)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e93 (35.36)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePancreas\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e248 (60.34)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e78 (52.70)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e170 (64.64)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBMI, kg/m2\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e22.43 (20.60, 24.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e22.48 (20.68, 24.75)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e22.39 (20.56, 24.18)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.305\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNRS-2002 score\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e2.00 (1.00, 4.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e3.00 (2.00, 4.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e2.00 (1.00, 4.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHemoglobin, g/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e123.00 (110.00, 134.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e125.00 (108.00, 135.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e123.00 (110.00, 134.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.895\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eWBC, 10^9/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e5.59 (4.66, 6.78)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e5.64 (4.76, 6.77)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5.50 (4.54, 6.78)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.255\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNeutrophil, 10^9/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e3.50 (2.70, 4.50)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e3.53 (2.81, 4.61)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e3.40 (2.67, 4.38)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.356\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLymphocyte, 10^9/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.40 (1.10, 1.79)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1.40 (1.09, 1.76)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e1.40 (1.13, 1.80)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.246\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePlatelet, 10^9/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e220.00 (174.50, 268.50)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e220.00 (175.50, 275.50)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e220.00 (174.50, 264.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.966\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMonocyte, 10^9/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.40 (0.30, 0.50)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.40 (0.30, 0.50)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.40 (0.30, 0.50)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.486\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAlbumin, g/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e37.00 (34.10, 40.48)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e36.80 (33.40, 40.20)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e37.30 (34.45, 40.60)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.381\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal bilirubin, \u0026micro;mol/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e19.55 (11.30, 72.38)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e21.70 (11.75, 96.55)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e18.00 (11.00, 64.35)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.036\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDirect bilirubin, \u0026micro;mol/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e4.80 (2.30, 35.80)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e6.20 (2.30, 51.75)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e4.45 (2.32, 31.85)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.058\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAST, U/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e32.00 (20.25, 76.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e33.50 (21.00, 71.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e30.00 (20.00, 76.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.694\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eALT, U/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e38.00 (17.00, 99.50)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e38.50 (17.00, 91.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e37.00 (17.00, 108.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.987\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTotal Cholesterol, mmol/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e4.61 (3.64, 5.43)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e4.71 (3.66, 5.27)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e4.57 (3.64, 5.52)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.725\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eHDL-C, mmol/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.97 (0.69, 1.19)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.90 (0.56, 1.13)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e1.00 (0.71, 1.19)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.026\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLDL-C, mmol/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e2.86 (2.17, 3.45)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e2.91 (2.18, 3.25)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e2.85 (2.17, 3.55)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.504\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCreatinine, \u0026micro;mol/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e66.60 (57.80, 78.70)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e70.30 (58.50, 83.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e65.40 (57.73, 75.45)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.030\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eGlucose, mmol/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e5.39 (4.80, 6.50)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e5.39 (4.80, 6.23)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e5.39 (4.81, 6.61)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.355\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eTriglyceride, mmol/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e1.33 (0.95, 1.83)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e1.33 (1.01, 1.97)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e1.34 (0.93, 1.80)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.178\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eUric Acid, \u0026micro;mol/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e265.00 (199.75, 337.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e276.00 (208.25, 341.50)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e262.00 (197.00, 330.75)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.213\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCA199, U/ml\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e42.65 (12.20, 164.20)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e46.25 (14.38, 163.90)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e40.30 (11.30, 163.90)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.693\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCA125, U/ml\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e12.20 (8.20, 19.60)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e12.75 (8.00, 20.95)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e12.10 (8.30, 18.20)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.497\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eCEA, \u0026micro;g/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e2.80 (1.80, 4.10)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e2.80 (1.87, 4.10)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e2.80 (1.80, 4.10)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.688\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePT, s\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e11.40 (10.90, 12.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e11.50 (10.90, 12.20)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e11.40 (10.90, 11.90)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.172\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAPTT, s\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e26.40 (25.10, 28.10)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e26.45 (25.20, 28.33)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e26.30 (24.95, 28.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.301\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eFibrinogen, g/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e3.38 (2.77, 4.08)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e3.40 (2.78, 4.19)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e3.35 (2.76, 4.04)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.798\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eD-dimer, \u0026micro;g/L\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e470.00 (260.00, 845.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e470.00 (250.00, 890.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e470.00 (270.00, 840.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.927\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eBlood loss, ml\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e200.00 (100.00, 400.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e200.00 (100.00, 500.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e200.00 (100.00, 300.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eOperation Time, min\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e385.00 (305.00, 450.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e405.00 (335.00, 477.50)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e370.00 (290.00, 420.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eApproach\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.002\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLPD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e203 (49.39)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e88 (59.46)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e115 (43.73)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRPD\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e208 (50.61)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e60 (40.54)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e148 (56.27)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eintraoperative RBC transfusion\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.013\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e276 (67.15)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e88 (59.46)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e188 (71.48)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e135 (32.85)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e60 (40.54)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e75 (28.52)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eConversion to open surgery\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c2\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c3\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"left\" colname=\"c4\"\u003e\u0026nbsp;\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.003\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNO\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e371 (90.27)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e125 (84.46)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e246 (93.54)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eYes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e40 (9.73)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e23 (15.54)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e17 (6.46)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"left\" colname=\"c5\"\u003e\u0026nbsp;\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003ePostoperative hospital stay, days\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e15.00 (11.00, 21.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e21.50 (15.00, 28.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e13.00 (11.00, 16.00)\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e\u0026lt;\u0026thinsp;.001\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"5\"\u003eAbbreviation: TO, textbook outcome; BMI, body mass index; NRS-2002 score, Nutritional Risk Screening 2002; WBC, white blood cell; AST, aspartate aminotransferase; ALT, alanine aminotransferase; HDL-C, high-density lipoprotein cholesterol; LDL-C, low-density lipoprotein cholesterol; CA19-9, carbohydrate antigen19-9; CEA, carcinoembryonic antigen; CA125, carbohydrate antigen125; PT, prothrombin time; APTT, activated partial thromboplastin time; RPD, robot-assisted pancreatoduodenectomy; LPD, laparoscopic pancreatoduodenectomy; RBC, red blood cell.\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003eThe TO group had a lower proportion of male patients (P\u0026thinsp;=\u0026thinsp;0.010), a higher proportion of lesions located in the pancreas (P\u0026thinsp;=\u0026thinsp;0.018), and a greater proportion of patients with a pancreatic duct diameter\u0026thinsp;\u0026gt;\u0026thinsp;3 mm (P\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The NRS-2002 score was significantly higher in the non-TO group (P\u0026thinsp;\u0026lt;\u0026thinsp;0.001).\u003c/p\u003e\u003cp\u003eRegarding laboratory measures, the non-TO group exhibited higher total bilirubin and creatinine levels (P\u0026thinsp;=\u0026thinsp;0.036 and P\u0026thinsp;=\u0026thinsp;0.030), whereas the TO group had significantly higher high-density lipoprotein cholesterol levels (P\u0026thinsp;=\u0026thinsp;0.026).\u003c/p\u003e\u003cp\u003eIntraoperative findings showed that the TO group experienced less intraoperative blood loss (P\u0026thinsp;\u0026lt;\u0026thinsp;0.001) and shorter operative times (P\u0026thinsp;\u0026lt;\u0026thinsp;0.001). The distribution of surgical approaches differed significantly between the groups, with a higher proportion of robotic surgeries in the TO group (P\u0026thinsp;=\u0026thinsp;0.002). Furthermore, the TO group had lower rates of intraoperative RBC transfusion (P\u0026thinsp;=\u0026thinsp;0.013) and conversion to open surgery (P\u0026thinsp;=\u0026thinsp;0.003).\u003c/p\u003e\u003c/div\u003e\n\u003ch3\u003eLASSO regression for feature selection\u003c/h3\u003e\n\u003cp\u003eA total of 411 patients were randomly divided into a training set (70%, n\u0026thinsp;=\u0026thinsp;287) and a validation set (30%, n\u0026thinsp;=\u0026thinsp;124) for model development and performance assessment (Supplementary Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). To mitigate the effects of collinearity, feature selection was performed using LASSO regression on the training set. By selecting a lambda (λ) value equal to one standard deviation from the minimum lambda, where the error is within one standard error of the minimum, eight variables were identified as the most predictive features (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e). The variables incorporated into the final model were: pancreatic duct diameter\u0026thinsp;\u0026gt;\u0026thinsp;3 mm, intraoperative blood loss, NRS-2002 score, direct bilirubin (DBIL), triglyceride (TG), carbohydrate antigen 125 (CA125), male, and BMI. These variables exhibited the strongest association with the outcome variable, while maintaining model simplicity and effectively reducing the risk of overfitting.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\n\u003ch3\u003eDevelop and evaluate the performance of the ML prediction models\u003c/h3\u003e\n\u003cp\u003eUsing the eight selected predictive variables, we developed and evaluated ML prediction models based on ten different algorithms. Ten classification algorithms were trained on the training set, including Logistic Regression, Random Forest, XGBoost, LightGBM, SVM, KNN, Decision Tree, MLP, AdaBoost, and Naive Bayes. Model performance was evaluated on the validation set using the AUC. Among all models, the Random Forest model achieved the highest discrimination, with an AUC of 0.86 on the validation set (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e). Additionally, the Brier score of the Random Forest model was 0.15, which was lower than those of several other models. Decision curve analysis (DCA) further demonstrated that the Random Forest model provided favorable net benefits for clinical decision-making across a wide range of threshold probabilities (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e). Therefore, the Random Forest model was identified as the optimal algorithm for predicting TO, outperforming other ML algorithms. Table\u0026nbsp;\u003cspan refid=\"Tab2\" class=\"InternalRef\"\u003e2\u003c/span\u003e summarizes the performance metrics of all models in the validation set.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv class=\"gridtable\"\u003e\u003ctable float=\"Yes\" id=\"Tab2\" border=\"1\"\u003e\u003ccaption language=\"En\"\u003e\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\u003cdiv class=\"CaptionContent\"\u003e\u003cp\u003eThe performance metrics of all models\u003c/p\u003e\u003c/div\u003e\u003c/caption\u003e\u003ccolgroup cols=\"8\"\u003e\u003cdiv align=\"left\" class=\"colspec\" colname=\"c1\" colnum=\"1\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c2\" colnum=\"2\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c3\" colnum=\"3\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c4\" colnum=\"4\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c5\" colnum=\"5\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c6\" colnum=\"6\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c7\" colnum=\"7\"\u003e\u003c/div\u003e\u003cdiv align=\"char\" char=\".\" class=\"colspec\" colname=\"c8\" colnum=\"8\"\u003e\u003c/div\u003e\u003cthead\u003e\u003ctr\u003e\u003cth align=\"left\" colname=\"c1\"\u003e\u003cp\u003eModel\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c2\"\u003e\u003cp\u003eAUC\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c3\"\u003e\u003cp\u003eBrier score\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c4\"\u003e\u003cp\u003eAccuracy\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c5\"\u003e\u003cp\u003eSensitivity\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c6\"\u003e\u003cp\u003eSpecificity\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c7\"\u003e\u003cp\u003ePPV\u003c/p\u003e\u003c/th\u003e\u003cth align=\"left\" colname=\"c8\"\u003e\u003cp\u003eNPV\u003c/p\u003e\u003c/th\u003e\u003c/tr\u003e\u003c/thead\u003e\u003ctbody\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eRandom Forest\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.86\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.15\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.80\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.77\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.84\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.90\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.68\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eXGBoost\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.82\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.72\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.65\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.84\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.88\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.58\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eSVM\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.83\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.76\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.77\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.73\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.84\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.65\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eKNN\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.78\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.69\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.67\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.71\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.80\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.55\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLightGBM\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.84\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.15\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.77\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.84\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.64\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.80\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.69\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eMLP\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.83\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.19\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.77\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.86\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.62\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.80\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.72\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eDecision Tree\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.74\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.21\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.74\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.76\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.71\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.82\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.63\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eLogistic\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.83\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.16\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.77\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.86\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.62\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.80\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.72\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eAdaBoost\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.78\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.18\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.73\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.75\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.71\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.82\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.62\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003ctr\u003e\u003ctd align=\"left\" colname=\"c1\"\u003e\u003cp\u003eNaive Bayes\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c2\"\u003e\u003cp\u003e0.83\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c3\"\u003e\u003cp\u003e0.19\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c4\"\u003e\u003cp\u003e0.75\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c5\"\u003e\u003cp\u003e0.68\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c6\"\u003e\u003cp\u003e0.87\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c7\"\u003e\u003cp\u003e0.90\u003c/p\u003e\u003c/td\u003e\u003ctd align=\"char\" char=\".\" colname=\"c8\"\u003e\u003cp\u003e0.61\u003c/p\u003e\u003c/td\u003e\u003c/tr\u003e\u003c/tbody\u003e\u003c/colgroup\u003e\u003ctfoot\u003e\u003ctr\u003e\u003ctd colspan=\"8\"\u003eAbbreviation: SVM, Support Vector Machine; KNN, K-Nearest Neighbors; LightGBM, Light Gradient Boosting Machine; MLP, Multi-Layer Perceptron; AdaBoost, Adaptive Boosting; AUC, area under the receiver operating characteristic curve; NPV, Negative predictive value; PPV, Positive predictive value\u003c/td\u003e\u003c/tr\u003e\u003c/tfoot\u003e\u003c/table\u003e\u003c/div\u003e\u003c/p\u003e\u003cdiv id=\"Sec11\" class=\"Section2\"\u003e\u003ch2\u003eModel interpretation\u003c/h2\u003e\u003cp\u003eThe SHAP summary plot illustrates the impact of each feature on the prediction model (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e). The included features were ranked according to their mean absolute SHAP values, from highest to lowest, reflecting their relative importance in the model. From the most to the least influential, the features were: pancreatic duct diameter\u0026thinsp;\u0026gt;\u0026thinsp;3 mm, intraoperative blood loss, NRS-2002 score, DBIL, TG, CA125, male, and BMI. According to the prediction model, the higher the SHAP value of the feature, the more likely it is to achieve TO. Finally, we developed an interactive interface that dynamically presents each patient\u0026rsquo;s predicted probability of achieving TO based on input variables. SHAP force plots visualize the main contributing factors and their specific effects, improving the model\u0026rsquo;s interpretability and clinical relevance (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eThis study systematically evaluated 40 potential predictors of TO and ultimately identified eight key variables: pancreatic duct diameter\u0026thinsp;\u0026gt;\u0026thinsp;3 mm, intraoperative blood loss, NRS-2002 score, DBIL, TG, CA125, male, and BMI. Based on these variables, ten ML models were developed to predict the achievement of TO after MIPD. Among them, the Random Forest model demonstrated superior predictive performance, with an AUC of 0.86. The advantage of the Random Forest model may be attributed to its strong ability to capture nonlinear relationships, model complex interactions among variables, and handle multicollinearity effectively. Furthermore, SHAP analysis was employed to interpret the prediction mechanism of the Random Forest model, providing insights into the relative contribution of each feature to the model\u0026rsquo;s decision-making process.\u003c/p\u003e\u003cp\u003eSurgical quality not only directly influences patients\u0026rsquo; short-term outcomes but is also closely associated with their long-term outcomes[\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. A single indicator is insufficient to comprehensively reflect the overall quality of a surgical procedure. The concept of the TO, as a comprehensive indicator, provides a more holistic and objective assessment of surgical quality[\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e]. This concept was first proposed in 2013 and applied to colorectal surgery, after which it was gradually extended to other surgical specialties[\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e14\u003c/span\u003e, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. In 2020, an international expert consensus further defined TO in pancreatic surgery[\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. With the widespread adoption of LPD and RPD worldwide, several studies have explored the factors associated with achieving TO after MIPD[\u003cspan additionalcitationids=\"CR17\" citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. However, no dedicated prediction model for TO after MIPD has yet been established. Therefore, this study aimed to develop a ML-based prediction model for TO following MIPD. The proposed model is intended to provide a scientific basis for preoperative risk assessment, surgical strategy optimization, and perioperative management, thereby improving surgical quality and enhancing patient outcomes.\u003c/p\u003e\u003cp\u003eSeveral key predictive features identified in this study have been verified in previous research. For example, Wu et al. reported that pancreatic duct dilation (diameter\u0026thinsp;\u0026gt;\u0026thinsp;3 mm) was associated with an increased probability of achieving TO after LPD[\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. Cai et al. showed that among patients with pancreatic duct dilation (diameter\u0026thinsp;\u0026gt;\u0026thinsp;3 mm), female patients were more likely than males to achieve TO after LPD[\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. Lee et al. suggested that a BMI\u0026thinsp;\u0026gt;\u0026thinsp;25 reduced the probability of achieving TO[\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. These findings further support the stability and reliability of the model in this study.\u003c/p\u003e\u003cp\u003eIn addition, this study identified several other important predictive features, including intraoperative blood loss, NRS-2002 score, DBIL, TG, CA125. The Shapley value plot showed that intraoperative blood loss was a key factor affecting the achievement of TO after MIPD, and its increase was closely related to a decrease in TO probability. Previous studies have confirmed that increased intraoperative blood loss is significantly associated with POPF, mortality, major complications, and readmission, all of which directly lead to TO failure[\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e, \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e]. It is necessary to perform preoperative nutritional assessment for all patients undergoing PD surgery[\u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e21\u003c/span\u003e]. The NRS-2002 score reflects patients\u0026rsquo; preoperative nutritional status, and a higher score indicates a greater risk of malnutrition. A meta-analysis by Sun et al. showed that higher NRS-2002 scores were significantly associated with increased postoperative complications and mortality in abdominal surgery[\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e], explaining the finding in this study that higher scores reduce the probability of achieving TO. It is worth noting that nutritional risk can be intervened through preoperative nutritional support. Xu et al. further confirmed that for high-risk patients undergoing OPD with NRS-2002\u0026thinsp;\u0026gt;\u0026thinsp;5, preoperative nutritional support can significantly reduce the incidence of POPF[\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. Two metabolism-related laboratory measures\u0026mdash;DBIL and TG\u0026mdash;also have predictive value for TO achievement. Elevated preoperative DBIL suggests biliary obstruction or impaired bile excretion, which can lead to hepatic dysfunction, systemic inflammatory response syndrome, or even neurotoxicity through accumulation in the central nervous system; these pathological processes may reduce the rate of TO achievement[\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e, \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. High preoperative TG levels are also associated with a lower TO achievement rate. Although direct evidence linking high TG levels to MIPD perioperative outcomes is still limited, existing studies in metabolism and cardiovascular fields suggest that elevated TG may be a risk marker for postoperative adverse outcomes[\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e, \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. CA125, a high-molecular-weight glycoprotein synthesized by mesothelial cells, can be elevated in various malignant tumors, acute inflammations, and certain physiological conditions[\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e, \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. CA125 levels reflect, to some extent, the tumor or inflammatory burden of the body. Recent studies have shown that CA125 can also serve as a potential marker of in-hospital mortality risk in patients with sepsis[\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e30\u003c/span\u003e]. Our study revealed a new association between CA125 and perioperative outcomes.\u003c/p\u003e\u003cp\u003eHowever, it is important to acknowledge several limitations of our study. First, as a retrospective study, it inherently carries selection bias, and all data were obtained from a single medical center. Second, the prediction model has only undergone internal validation and has not yet been tested in external cohorts, which to some extent weakens the reliability of our findings. In the future, we plan to conduct prospective studies and multi-center collaborative research to further validate and optimize the model in order to achieve more accurate results. We will also continue to report our research findings.\u003c/p\u003e"},{"header":"CONCLUSIONS","content":"\u003cp\u003eOur study introduced a new machine learning algorithm to predict the probability of achieving TO after MIPD, which demonstrated excellent performance. This model may assist in preoperative risk assessment, surgical strategy optimization, and perioperative management for patients undergoing MIPD, although external validation is still required before it can be widely applied in clinical practice.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003ch2\u003eCompeting interests\u003c/h2\u003e\u003cp\u003eThe authors have no conflicts of interest to declare.\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e\u003cp\u003eThis study was supported by the National Science and Technology Major Project (2025ZD0552312), Zhejiang Provincial Medical and Health Technology Program (No. 2023KY517), and Key Project of social welfare program of Zhejiang Science and Technology Department, \u0026ldquo;Lingyan\u0026rdquo;Program (2022C03099).\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003ePc.M. , Writing \u0026ndash; review \u0026amp; editing, Writing \u0026ndash; original draft, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization; Zc.J. , Writing \u0026ndash; review \u0026amp; editing, Writing \u0026ndash; original draft, Resources, Project administration, Methodology, Funding acquisition, Data curation; Yy.W. , Investigation, Formal analysis, Data curation; Z.J. , Data curation; Za.Z. , Data curation; Yp.M. , Writing \u0026ndash; review \u0026amp; editing, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization; Ww. J. , Writing \u0026ndash; review \u0026amp; editing, Visualization, Validation, Supervision, Software, Resources, Project administration, Methodology, Investigation, Funding acquisition, Formal analysis, Data curation, Conceptualization.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eDeidentified data and the R code for model building and validation will be made available from the corresponding author upon reasonable request after publication, subject to institutional data-sharing policies.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eZhang XP, Xu S, Zhao ZM, Yu GS, Han B, Chen X, Ma YT, Xu ZZ, Liu Z, Li ES, Guo XF, Gao YX, Zhao GD, Lau WY, Liu J, Liu R (2023) Outcomes of Robotic Versus Laparoscopic Pancreatoduodenectomy Following Learning Curves of Surgeons: A Multicenter Study on 2255 Patients. Ann Surg\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang XJ, Fei H, Guo CG, Sun CY, Li ZF, Li Z, Chen YT, Che X, Zhao DB (2023) Analysis of textbook outcomes for ampullary carcinoma patients following pancreaticoduodenectomy. World J Gastrointest Surg 15:2259\u0026ndash;2271\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKolfschoten NE, Kievit J, Gooiker GA, van Leersum NJ, Snijders HS, Eddes EH, Tollenaar RA, Wouters MW, -van de Marang PJ (2013) Focusing on desired outcomes of care after colon cancer resections; hospital variations in 'textbook outcome'. Eur J Surg Oncol 39:156\u0026ndash;163\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003evan Roessel S, Mackay TM, van Dieren S, van der Schelling GP, Nieuwenhuijs VB, Bosscha K, van der Harst E, van Dam RM, Liem MSL, Festen S, Stommel MWJ, Roos D, Wit F, Molenaar IQ, de Meijer VE, Kazemier G, de Hingh I, van Santvoort HC, Bonsing BA, Busch OR, Groot Koerkamp B, Besselink MG (2020) Textbook Outcome: Nationwide Analysis of a Novel Quality Measure in Pancreatic Surgery. Ann Surg 271:155\u0026ndash;162\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eYuan J, Du C, Wu H, Zhong T, Zhai Q, Peng J, Liu N, Li J (2025) Risk factors of failure to achieve textbook outcome in patients after pancreatoduodenectomy: a systematic review and meta-analysis. Int J Surg 111:3093\u0026ndash;3106\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eZhang Z, Zhao X, Shang M, Xu Q, Wang X, Zhang J, Wang C, Gu Z (2025) Machine learning model for postpancreaticoduodenectomy haemorrhage prediction: an international multicentre cohort study. BMJ Open 15:e096147\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXu J, Chen T, Fang X, Xia L, Pan X (2024) Prediction model of pressure injury occurrence in diabetic patients during ICU hospitalization\u0026ndash;XGBoost machine learning model can be interpreted based on SHAP. Intensive Crit Care Nurs 83:103715\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKondrup J, Rasmussen HH, Hamberg O, Stanga Z (2003) Nutritional risk screening (NRS 2002): a new method based on an analysis of controlled clinical trials. Clin Nutr 22:321\u0026ndash;336\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eClavien PA, Barkun J, de Oliveira ML, Vauthey JN, Dindo D, Schulick RD, de Santiba\u0026ntilde;es E, Pekolj J, Slankamenac K, Bassi C, Graf R, Vonlanthen R, Padbury R, Cameron JL, Makuuchi M (2009) The Clavien-Dindo classification of surgical complications: five-year experience. Ann Surg 250:187\u0026ndash;196\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBassi C, Marchegiani G, Dervenis C, Sarr M, Abu Hilal M, Adham M, Allen P, Andersson R, Asbun HJ, Besselink MG, Conlon K, Del Chiaro M, Falconi M, Fernandez-Cruz L, Fernandez-Del Castillo C, Fingerhut A, Friess H, Gouma DJ, Hackert T, Izbicki J, Lillemoe KD, Neoptolemos JP, Olah A, Schulick R, Shrikhande SV, Takada T, Takaori K, Traverso W, Vollmer CM, Wolfgang CL, Yeo CJ, Salvia R, Buchler M (2017) The 2016 update of the International Study Group (ISGPS) definition and grading of postoperative pancreatic fistula: 11 Years After. Surgery 161:584\u0026ndash;591\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWente MN, Veit JA, Bassi C, Dervenis C, Fingerhut A, Gouma DJ, Izbicki JR, Neoptolemos JP, Padbury RT, Sarr MG, Yeo CJ, B\u0026uuml;chler MW (2007) Postpancreatectomy hemorrhage (PPH): an International Study Group of Pancreatic Surgery (ISGPS) definition. Surgery 142:20\u0026ndash;25\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKoch M, Garden OJ, Padbury R, Rahbari NN, Adam R, Capussotti L, Fan ST, Yokoyama Y, Crawford M, Makuuchi M, Christophi C, Banting S, Brooke-Smith M, Usatoff V, Nagino M, Maddern G, Hugh TJ, Vauthey JN, Greig P, Rees M, Nimura Y, Figueras J, DeMatteo RP, B\u0026uuml;chler MW, Weitz J (2011) Bile leakage after hepatobiliary and pancreatic surgery: a definition and grading of severity by the International Study Group of Liver Surgery. Surgery 149:680\u0026ndash;688\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWang H, Hu X, Yin C, Zhou D, Li Z, Ma Z, Zhang H (2024) Association of textbook outcomes with improved survival in pancreatic ductal adenocarcinoma following pancreaticoduodenectomy: a retrospective study. Transl Gastroenterol Hepatol 9:38\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLim C, Llado L, Salloum C, Ramos E, Lopez-Dominguez J, Cachero A, Fabregat J, Azoulay D (2021) Textbook Outcome Following Liver Transplantation. World J Surg 45:3414\u0026ndash;3423\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWu MY, McGregor RJ, Scott J, Smithers BM, Thomas J, Frankel A, Barbour A, Thomson I (2023) Textbook outcomes for oesophagectomy: A valid composite measure assessment tool for surgical performance in a specialist unit. Eur J Surg Oncol 49:106897\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCai H, Lu F, Gao P, Zhang M, Wang X, Li Y, Meng L, Peng B, Cai Y (2024) Risk factors of textbook outcome in laparoscopic pancreatoduodenectomy: results from a prospective high-volume center study. BMC Surg 24:233\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWu Y, Peng B, Liu J, Yin X, Tan Z, Liu R, Hong D, Zhao W, Wu H, Chen R, Li D, Huang H, Miao Y, Liu Y, Liang T, Wang W, Yuan J, Li S, Zhang H, Wang M, Qin R (2023) Textbook outcome as a composite outcome measure in laparoscopic pancreaticoduodenectomy: a multicenter retrospective cohort study. Int J Surg 109:374\u0026ndash;382\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLee B, Han HS, Yoon YS, Lee JS (2025) Textbook Outcomes of Totally Robotic Versus Totally Laparoscopic Pancreaticoduodenectomy for Periampullary Neoplasm: A Propensity Score-Matched Cohort Study. J Clin Med 14\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSeykora TF, Ecker BL, McMillan MT, Maggino L, Beane JD, Fong ZV, Hollis RH, Jamieson NB, Javed AA, Kowalsky SJ, Kunstman JW, Malleo G, Poruk KE, Soares K, Valero V 3rd, Velu LKP, Watkins AA, Vollmer CM Jr (2019) The Beneficial Effects of Minimizing Blood Loss in Pancreatoduodenectomy. Ann Surg 270:147\u0026ndash;157\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCasciani F, Trudeau MT, Asbun HJ, Ball CG, Bassi C, Behrman SW, Berger AC, Bloomston MP, Callery MP, Christein JD, Falconi M, Fernandez-Del Castillo C, Dillhoff ME, Dickson EJ, Dixon E, Fisher WE, House MG, Hughes SJ, Kent TS, Kunstman JW, Malleo G, Partelli S, Wolfgang CL, Zureikat AH, Vollmer CM (2021) The effect of high intraoperative blood loss on pancreatic fistula development after pancreatoduodenectomy: An international, multi-institutional propensity score matched analysis. Surgery 170:1195\u0026ndash;1204\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAleassa EM, Morris-Stiff G (2019) Regarding: Nutritional support and therapy in pancreatic surgery: A position paper of the International Study Group on Pancreatic Surgery (ISGPS). Surgery 165:1248\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSun Z, Kong XJ, Jing X, Deng RJ, Tian ZB (2015) Nutritional Risk Screening 2002 as a Predictor of Postoperative Outcomes in Patients Undergoing Abdominal Surgery: A Systematic Review and Meta-Analysis of Prospective Cohort Studies. PLoS ONE 10:e0132857\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eXu JY, Tian XD, Song JH, Chen J, Yang YM, Wei JM (2021) Preoperative Nutrition Support May Reduce the Prevalence of Postoperative Pancreatic Fistula after Open Pancreaticoduodenectomy in Patients with High Nutritional Risk Determined by NRS2002. Biomed Res Int 2021:6691966\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLiu JJ, Sun YM, Xu Y, Mei HW, Guo W, Li ZL (2023) Pathophysiological consequences and treatment strategy of obstructive jaundice. World J Gastrointest Surg 15:1262\u0026ndash;1276\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePavlidis ET, Pavlidis TE (2018) Pathophysiological consequences of obstructive jaundice and perioperative management. Hepatobiliary Pancreat Dis Int 17:17\u0026ndash;21\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eChen C, Wen Q, Ma C, Li X, Huang T, Ke J, Gong C, Hei Z (2022) Hypertriglyceridemia is associated with stroke after non-cardiac, non-neurological surgery in the older patients: A nested case-control study. Front Aging Neurosci 14:935934\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMurthy NM, Yoo TT, Sanchez A, Chhitu M, Abramov D, Gatling J, Mamas MA, Parwani P (2025) Preoperative Diagnostic Assessment of Patients with Cardiovascular Risk Factors Undergoing Noncardiac Surgery: A 2025 Update. Methodist Debakey Cardiovasc J 21:87\u0026ndash;100\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBuamah P (2000) Benign conditions associated with raised serum CA-125 concentration. J Surg Oncol 75:264\u0026ndash;265\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSj\u0026ouml;vall K, Nilsson B, Einhorn N (2002) The significance of serum CA 125 elevation in malignant and nonmalignant diseases. Gynecol Oncol 85:175\u0026ndash;178\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGomar S, Tejeda L, Bou R, Romero B, Quesada-Dorador A (2022) Association of carbohydrate 125 antigen with sepsis mortality in critical patients. Med Clin (Barc) 159:124\u0026ndash;129\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"Minimally invasive pancreaticoduodenectomy, Textbook outcome, Machine learning, Random Forest, Surgical quality","lastPublishedDoi":"10.21203/rs.3.rs-8269976/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8269976/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eBackground:\u003c/h2\u003e\u003cp\u003eTextbook Outcome (TO) reflects overall surgical quality. With the expanding use of minimally invasive pancreaticoduodenectomy (MIPD), reliable prediction of TO is essential. This study aimed to identify predictors of TO after MIPD and develop a machine learning (ML) model.\u003c/p\u003e\u003ch2\u003eMethods:\u003c/h2\u003e\u003cp\u003eWe retrospectively analyzed 411 patients undergoing MIPD (2017\u0026ndash;2023). The Least Absolute Shrinkage and Selection Operator regression was used for feature selection, and ten ML algorithms were compared. Model performance was evaluated using the area under the receiver operating characteristic curve (AUC). Model interpretability was assessed using SHapley Additive exPlanations (SHAP).\u003c/p\u003e\u003ch2\u003eResults:\u003c/h2\u003e\u003cp\u003eOf 411 patients, 263 (63.99%) achieved TO. Eight variables were identified as predictive features. Among the ten algorithms, the Random Forest model demonstrated the best discrimination (AUC\u0026thinsp;=\u0026thinsp;0.86).\u003c/p\u003e\u003ch2\u003eConclusions:\u003c/h2\u003e\u003cp\u003eThe Random Forest model accurately predicted TO after MIPD and may assist in individualized preoperative risk stratification and perioperative management.\u003c/p\u003e","manuscriptTitle":"Development and Validation of a Machine Learning–Based Model for Predicting Textbook Outcome after Minimally Invasive Pancreaticoduodenectomy","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-10 12:44:23","doi":"10.21203/rs.3.rs-8269976/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"2cd8c493-c4ce-4e05-b6f7-33d8b160a352","owner":[],"postedDate":"December 10th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-01-29T15:53:18+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-10 12:44:23","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8269976","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8269976","identity":"rs-8269976","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}