2.Application of machine learning model based on XGBoost algorithm in early prediction of patients with acute severe pancreatitis.
Xin GAO ; Jiaxi LIN ; Airong WU ; Huiyuan GU ; Xiaolin LIU ; Minyue YIN ; Zhirun ZHOU ; Rufa ZHANG ; Chunfang XU ; Jinzhou ZHU
Chinese Critical Care Medicine 2023;35(4):421-426
OBJECTIVE:
To establish a machine learning model based on extreme gradient boosting (XGBoost) algorithm for early prediction of severe acute pancreatitis (SAP), and explore its predictive efficiency.
METHODS:
A retrospective cohort study was conducted. The patients with acute pancreatitis (AP) who admitted to the First Affiliated Hospital of Soochow University, the Second Affiliated Hospital of Soochow University and Changshu Hospital Affiliated to Soochow University from January 1, 2020 to December 31, 2021 were enrolled. Demography information, etiology, past history, and clinical indicators and imaging data within 48 hours of admission were collected according to the medical record system and image system, and the modified CT severity index (MCTSI), Ranson score, bedside index for severity in acute pancreatitis (BISAP) and acute pancreatitis risk score (SABP) were calculated. The data sets of the First Affiliated Hospital of Soochow University and Changshu Hospital Affiliated to Soochow University were randomly divided into training set and validation set according to 8 : 2. Based on XGBoost algorithm, the SAP prediction model was constructed on the basis of hyperparameter adjustment by 5-fold cross validation and loss function. The data set of the Second Affiliated Hospital of Soochow University was served as independent test set. The predictive efficacy of the XGBoost model was evaluated by drawing the receiver operator characteristic curve (ROC curve), and compared it with the traditional AP related severity score; variable importance ranking diagram and Shapley additive explanation (SHAP) diagram were drawn to visually explain the model.
RESULTS:
A total of 1 183 AP patients were enrolled finally, of which 129 (10.9%) developed SAP. Among the patients from the First Affiliated Hospital of Soochow University and Changshu Hospital Affiliated to Soochow University, there were 786 patients in the training set and 197 in the validation set; 200 patients from the Second Affiliated Hospital of Soochow University were used as the test set. Analysis of all three datasets showed that patients who advanced to SAP exhibited pathological manifestation such as abnormal respiratory function, coagulation function, liver and kidney function, and lipid metabolism. Based on the XGBoost algorithm, an SAP prediction model was constructed, and ROC curve analysis showed that the accuracy for prediction of SAP reached 0.830, the area under the ROC curve (AUC) was 0.927, which was significantly improved compared with the traditional scoring systems including MCTSI, Ranson, BISAP and SABP, the accuracy was 0.610, 0.690, 0.763, 0.625, and the AUC was 0.689, 0.631, 0.875, and 0.770, respectively. The feature importance analysis based on the XGBoost model showed that the top ten items ranked by the importance of model features were admission pleural effusion (0.119), albumin (Alb, 0.049), triglycerides (TG, 0.036), Ca2+ (0.034), prothrombin time (PT, 0.031), systemic inflammatory response syndrome (SIRS, 0.031), C-reactive protein (CRP, 0.031), platelet count (PLT, 0.030), lactate dehydrogenase (LDH, 0.029), and alkaline phosphatase (ALP, 0.028). The above indicators were of great significance for the XGBoost model to predict SAP. The SHAP contribution analysis based on the XGBoost model showed that the risk of SAP increased significantly when patients had pleural effusion and decreased Alb.
CONCLUSIONS
A SAP prediction scoring system was established based on the machine automatic learning XGBoost algorithm, which can predict the SAP risk of patients within 48 hours of admission with good accuracy.
Humans
;
Pancreatitis
;
Acute Disease
;
Retrospective Studies
;
Hospitalization
;
Algorithms
3.Research progress of hypertonic saline in early resuscitation of acute pancreatitis.
Chinese Critical Care Medicine 2023;35(4):446-448
Local inflammatory reaction and microcirculation disturbance are the early manifestations of acute pancreatitis (AP). Studies have shown that early and reasonable fluid resuscitation of patients with AP can reduce related complications and prevent the deterioration to severe acute pancreatitis (SAP). Traditional isotonic crystalloid (such as Ringer solution) is considered to be a safe and reliable resuscitation solution, but too much and too fast infusion in the early stage of shock will increase the risk of complications such as tissue edema and abdominal compartment syndrome (ACS). Many scholars have found that hypertonic saline resuscitation solution has the advantages of reducing tissue and organ edema, rapidly restoring hemodynamics, inhibiting oxidative stress and inflammatory signal transduction, thereby improving the prognosis of AP patients and reducing the incidence of SAP and mortality. This article summarizes the mechanisms of hypertonic saline in the resuscitation treatment of AP patients in recent years, in order to provide reference for the clinical application and research of AP patients.
Humans
;
Pancreatitis
;
Acute Disease
;
Resuscitation
;
Inflammation
;
Crystalloid Solutions
;
Saline Solution, Hypertonic
4.Effect of early fluid balance on the prognosis in severe acute pancreatitis.
Cong ZHANG ; Yalin OU ; Hongliang QIAN ; Yinxia XU
Chinese Critical Care Medicine 2023;35(5):524-527
OBJECTIVE:
To observe the correlation between early fluid resuscitation and prognosis in patients with severe acute pancreatitis (SAP).
METHODS:
SAP patients admitted to the department of critical care medicine of the People's Hospital of Chuxiong Yi Autonomous Prefecture of Yunnan Province from June 2018 to December 2020 were enrolled and analyzed retrospectively. All patients were given the routine treatment according to their condition and relevant diagnostic According to their different prognosis, enrolled patients were divided into death group and survival group. The differences in gender, age, acute physiology and chronic health evaluation II (APACHE II) and Ranson score on admission between the two groups were analyzed. Taking 24 hours as an observation day, the fluid inflow, outflow, and net balance at the first, second, and third 24 hours after admission were recorded, and the ratio of the fluid inflow at the first 24 hours to the total fluid inflow in 72 hours (FV24 h-1 st) was calculated as a study index. Using 33% as the standard, compare the proportion of patients in the two groups who achieved FV24 h-1 st < 33%. The differences of various indicators between the two groups were compared, and the effect of early fluid balance on the prognosis of SAP patients was analyzed.
RESULTS:
Eighty-nine patients were included in the study (41 in the death group, 48 in the survival group). There were no statistically significant differences on age (years old: 57.6±15.2 vs. 49.5±15.2), gender (male: 61.0% vs. 54.2%), APACHE II score (18.0±2.4 vs. 17.3±2.3), and Ranson score (6.3±1.4 vs. 5.9±1.2) between the death group and the survival group at the time of admission on the intensive care unit (ICU) (all P > 0.05). The fluid intake of the death group in the first 24 hours, the second 24 hours and the third 24 hours after admission to ICU was significantly higher than that of the survival group, and the difference was statistically significant (mL: 4 138±832 vs. 3 535±1 058, 3 883±729 vs. 3 324±516, 3 786±490 vs. 3 212±609, all P < 0.05), and the fluid inflow in the death group at the first 24 hours was greater than 4 100 mL. After treatment, the fluid outflow of the death group at the three 24-hour periods after admission on the ICU was an increasing trend, but it was still significantly less than that of the survival group at the three 24-hour periods (mL: 1 242±465 vs. 1 795±819, 1 536±579 vs. 2 080±524, 1 610±585 vs. 2 932±752, all P < 0.01). Due to the fact that the total fluid inflow and total fluid outflow in the three 24-hour periods in the death group were more than those in the survival group, the net fluid balances in the three 24-hour periods in the death group were still significantly more than those in the survival group finally (mL: 2 896±782 vs. 1 740±725, 2 347±459 vs. 1 243±795, 2 176±807 vs. 338±289, all P < 0.01). There was no difference in FV24 h-1 st between the death group and survival group [FV24 h-1 st > 33%: 56.1% (23/41) vs. 54.2% (26/48), P > 0.05].
CONCLUSIONS
Fluid resuscitation is an important method for early treatment of SAP, but it also has many adverse reactions. Fluid resuscitation indexes such as fluid inflow, outflow, net balance, and FV24 h-1 st within 24 to 72 hours after admission are related to the prognosis of patients with SAP, and can be used as indicators to evaluate the prognosis of SAP. The optimized fluid resuscitation strategy can improve the prognosis of patients with SAP.
Humans
;
Male
;
Acute Disease
;
Retrospective Studies
;
Pancreatitis
;
China
;
Prognosis
;
Water-Electrolyte Balance
5.Evolution and progress of surgical intervention strategies for acute pancreatitis.
Rui BAI ; Tian Qi LU ; Bei SUN
Chinese Journal of Surgery 2023;61(7):556-561
Acute pancreatitis is one of the common surgical acute abdominal diseases. Since people first recognized acute pancreatitis in the middle of the nineteenth century, a diversified minimally invasive treatment model with standardization has been formed today. According to the main line of surgical intervention of acute pancreatitis treatment,this period can be roughly divided into five stages:exploration stage, conservative treatment stage, pancreatectomy stage, debridement and drainage of the pancreatic necrotic tissue stage, and minimally invasive treatment as the first choice led by the multidisciplinary team mode stage. Throughout history, the evolution and progress of surgical intervention strategies for acute pancreatitis cannot be separated from the progress of science and technology, the update of treatment concepts and the further understanding of the pathogenesis. This article will summarize the surgical characteristics of acute pancreatitis treatment at each stage to explain the development of surgical treatment of acute pancreatitis,to help investigate the development of surgical treatment of acute pancreatitis in the future.
Humans
;
Acute Disease
;
Treatment Outcome
;
Debridement
;
Pancreatitis, Acute Necrotizing/pathology*
;
Drainage
6.Clinical characteristics of patients with colon complications after necrotizing pancreatitis: a retrospective cohort study.
Feng CAO ; Jiong Di LU ; Ang LI ; Chao ZHANG ; Zhe WANG ; Chong Chong GAO ; Xiao Hui WANG ; Fei LI
Chinese Journal of Surgery 2023;61(7):567-574
Objective: To investigate the clinical characteristics of colon complications in patients with necrotizing pancreatitis(NP). Methods: The clinical data of 403 patients with NP admitted to the Department of General Surgery,Xuanwu Hospital, Capital Medical University from January 2014 to December 2021 were retrospectively analyzed. There were 273 males and 130 females,aged (49.4±15.4) years(range: 18 to 90 years). Among them,there were 199 cases of biliary pancreatitis,110 cases of hyperlipidemic pancreatitis,and 94 cases of pancreatitis caused by other causes. A multidisciplinary diagnosis and treatment model was used to diagnose and treat patients. Depending on whether the patients had colon complications,they were divided into colon complications group and noncolon complications group. Patients with colon complications were treated with anti-infection therapy,parental nutritional support,keeping the drainage tube unobstructed,and terminal ileostomy. The clinical results of the two groups were compared and analyzed using a 1∶1 propensity score match(PSM) method. The t test,χ2 test, or rank-sum test was used to analyze data between groups,respectively. Results: The incidence of colon complications was 13.2%(53/403),including 15 cases of colon obstruction,23 cases of colon fistula,and 21 cases of colon hemorrhage. After PSM,the baseline and clinical characteristics at admission of the two groups of patients were comparable (all P>0.05). In terms of clinical outcome,compared to patients with NP without colon complications,the number of patients with colon complications who received minimally invasive intervention(88.7%(47/53) vs. 69.8%(37/53),χ2=5.736,P=0.030),the number of minimally invasive interventions (M(IQR))(2(2) vs. 1(1), Z=4.638,P=0.034),the number of patients with multiple organ failure(45.3%(24/53) vs. 32.1%(17/53),χ2=4.826,P=0.041),and the number of extrapancreatic infections(79.2%(42/53) vs. 60.4%(32/53),χ2=4.476,P=0.034) increased significantly. The time required for enteral nutrition support(8(30)days vs. 2(10) days, Z=-3.048, P=0.002), parental nutritional support(32(37)days vs. 17(19)days, Z=-2.592, P=0.009),the length of stay in the ICU(24(51)days vs. 18(31)days, Z=-2.268, P=0.002),and the total length of stay (43(52)days vs. 30(40)days, Z=-2.589, P=0.013) were also significantly prolonged. However,mortality rates in the two groups were similar(37.7%(20/53) vs. 34.0%(18/53),χ2=0.164,P=0.840). Conclusions: Colonic complications in NP patients are not rare,which can lead to prolonged hospitalization and increased surgical intervention. Active surgical intervention can help improve the prognosis of these patients.
Male
;
Female
;
Humans
;
Retrospective Studies
;
Pancreatitis, Acute Necrotizing/surgery*
;
Prognosis
;
Colon
;
Treatment Outcome
7.Effect of pancreatic extracorporeal shock wave lithotripsy on chronic pancreatitis stones.
Wan Xing DUAN ; Wan Zhen WEI ; Xue YANG ; Qi GAO ; Jun CHEN ; Zheng WU ; Zheng WANG
Chinese Journal of Surgery 2023;61(7):590-595
Objective: To analyze the therapeutic effect and safety of pancreatic extracorporeal shock wave lithotripsy(P-ESWL) for patients with chronic pancreatitis complicated by stones of the pancreatic duct and to investigate the influencing factors. Methods: A retrospective analysis was performed on clinical data from 81 patients with chronic pancreatitis complicated by pancreatic duct calculus treated with P-ESWL in the Department of Hepatobiliary Surgery, the First Affiliated Hospital of Xi 'an Jiaotong University from July 2019 to May 2022. There were 55 males(67.9%) and 26 females(32.1%). The age was (47±15)years (range: 17 to 77 years). The maximum diameter(M(IQR)) of the stone was 11.64(7.60) mm, and the CT value of the stone was 869 (571) HU. There were 32 patients (39.5%) with a single pancreatic duct stone and 49 patients(60.5%) with multiple pancreatic duct stones. The effectiveness, remission rate of abdominal pain, and complications of P-ESWL were evaluated. Student's t test, Mann Whitney U test, χ2 test, or Fisher's exact test was used to compare the characteristics between the effective and ineffective groups of lithotripsy. The factors influencing the effect of lithotripsy were analyzed by univariate and multivariate logistic regression analysis. Results: Eighty-one patients with chronic pancreatitis were treated with P-ESWL 144 times, with an average of 1.78 (95%CI:1.60 to 1.96) times per person. Among them, 38 patients(46.9%) were treated with endoscopy. There were 64 cases(79.0%) with effective removal of pancreatic duct calculi and 17 cases(21.0%) with ineffective removal. Of the 61 patients with chronic pancreatitis accompanied by abdominal pain, 52 cases(85.2%) had pain relief after lithotripsy. After lithotripsy treatment, 45 patients(55.6%) developed skin ecchymosis, 23 patients(28.4%) had sinus bradycardia, 3 patients(3.7%) had acute pancreatitis, 1 patient(1.2%) had a stone lesion, and 1 patient(1.2%) had a hepatic hematoma. Univariate and multivariate logistic regression analysis showed that the factors affecting the efficacy of lithotripsy included the age of patient(OR=0.92, 95%CI: 0.86 to 0.97), the maximum diameter of the stone(OR=1.12,95%CI:1.02 to 1.24) and the CT value of the stone(OR=1.44, 95%CI: 1.17 to 1.86). Conclusions: P-ESWL is effective in the treatment of patients with chronic pancreatitis complicated by calculi of the main pancreatic duct.Factors affecting the efficacy of lithotripsy include patient's age, maximum stone diameter, and CT value of calculi.
Male
;
Female
;
Humans
;
Retrospective Studies
;
Acute Disease
;
Treatment Outcome
;
Calculi/pathology*
;
Lithotripsy
;
Pancreatitis, Chronic/pathology*
;
Pancreatic Diseases/complications*
;
Pancreatic Ducts
;
Abdominal Pain/therapy*
8.Advances in the study of postpancreatectomy acute pancreatitis.
Chinese Journal of Surgery 2023;61(7):609-613
Postpancreatectomy acute pancreatitis (PPAP) is an acute inflammation of the remnant pancreas in the early postoperative period caused by a variety of factors. With the progress of related research,PPAP has been confirmed as an independent risk factor for many severe complications such as postoperative pancreatic fistula. In some cases, it progresses to necrotizing PPAP, increasing the risk of mortality. Currently, the International Study Group for Pancreatic Surgery has standardized and graded PPAP as an independent complication, taking into account factors including serum amylase, radiological features, and clinical impact. This review summarizes how the concept of PPAP was proposed, as well as the latest progress in the research related to its etiology, prognosis, prevention, and treatment. However, given the large heterogeneity of relevant studies and the fact that they were mostly retrospective, in the future, it is necessary to place more emphasis on PPAP and elucidate the problems through more standardized studies to optimize strategies for the prevention and management of complications after pancreatic surgery.
Humans
;
Pancreatitis/complications*
;
Retrospective Studies
;
Acute Disease
;
Pancreas
;
Postoperative Complications/etiology*
;
Pancreatic Fistula/etiology*
;
Pancreaticoduodenectomy/adverse effects*
9.Basic principles,methods and evaluation of minimally invasive treatment for infected pancreatic necrosis.
Chinese Journal of Surgery 2023;61(1):13-17
Infected pancreatic necrosis(IPN) is the main surgical indication of acute pancreatitis. Minimally invasive debridement has become the mainstream surgical strategy of IPN,and it is only preserved for IPN patients who are not response for adequate non-surgical treatment. Transluminal or retroperitoneal drainage is preferred,and appropriate debridement can be performed. At present,it is reported that video assisted transluminal,trans-abdominal and retroperitoneal approaches can effectively control IPN infection. However,in terms of reducing pancreatic leakage and other complications,surgical and endoscopic transgastric debridement may be the future direction in the treatment of IPN.
Humans
;
Pancreatitis, Acute Necrotizing/complications*
;
Acute Disease
;
Debridement/methods*
;
Endoscopy/methods*
;
Drainage/methods*
;
Intraabdominal Infections/complications*
;
Treatment Outcome
10.Imaging classification and analysis of the diagnosis and treatment of infected pancreatic necrosis:a report of 126 cases.
Tian Qi LU ; Li Ren SHANG ; Fan BIE ; Yi Lin XU ; Yu Hang SUI ; Guan Qun LI ; Hua CHEN ; Gang WANG ; Rui KONG ; Xue Wei BAI ; Hong Tao TAN ; Yong Wei WANG ; Bei SUN
Chinese Journal of Surgery 2023;61(1):33-40
Objective: To explore the clinical characteristics of various types of infected pancreatic necrosis(IPN) and the prognosis of different treatment methods in the imaging classification of IPN proposed. Methods: The clinical data of 126 patients with IPN admitted to the Department of Pancreatic and Biliary Surgery, the First Affiliated Hospital of Harbin Medical University from December 2018 to December 2021 were analyzed retrospectively. There were 70 males(55.6%) and 56 females(44.4%), with age(M(IQR)) of 44(17)years (range: 12 to 87 years). There were 67 cases(53.2%) of severe acute pancreatitis and 59 cases (46.8%) of moderately severe acute pancreatitis. All cases were based on the diagnostic criteria of IPN. All cases were divided into Type Ⅰ(central IPN)(n=21), Type Ⅱ(peripheral IPN)(n=23), Type Ⅲ(mixed IPN)(n=74) and Type Ⅳ(isolated IPN)(n=8) according to the different sites of infection and necrosis on CT.According to different treatment strategies,they were divided into Step-up group(n=109) and Step-jump group(n=17). The clinical indicators and prognosis of each group were observed and analyzed by ANOVA,t-test,χ2 test or Fisher exact test,respectively. Results: There was no significant difference in mortality, complication rate and complication grade in each type of IPN(all P>0.05). Compared with other types of patients, the length of stay (69(40)days vs. 19(19)days) and hospitalization expenses(323 000(419 000)yuan vs. 60 000(78 000)yuan) were significantly increased in Type Ⅳ IPN(Z=-4.041, -3.972; both P<0.01). The incidence of postoperative residual infection of Type Ⅳ IPN was significantly higher than that of other types (χ2=16.350,P<0.01). There was no significant difference in the mortality of patients with different types of IPN between different treatment groups. The length of stay and hospitalization expenses of patients in the Step-up group were significantly less than those in the Step-jump group(19(20)days vs. 33(35)days, Z=-2.052, P=0.040;59 000(80 000)yuan vs. 122 000(109 000)yuan,Z=-2.317,P=0.020). Among the patients in Type Ⅳ IPN, the hospitalization expenses of Step-up group was significantly higher than that of Step-jump group(330 000(578 000)yuan vs. 141 000 yuan,Z=-2.000,P=0.046). The incidence of postoperative residual infection of Step-up group(17.4%(19/109)) was significantly lower than that of Step-jump group(10/17)(χ2=11.980, P=0.001). Conclusions: Type Ⅳ IPN is more serious than the other three types. It causes longer length of stay and more hospitalization expenses. The step-up approach is safe and effective in the treatment of IPN. However, for infected lesions which are deep in place,difficult to reach by conventional drainage methods, or mainly exhibit "dry necrosis", choosing the step-jump approach is a more positive choice.
Male
;
Female
;
Humans
;
Retrospective Studies
;
Pancreatitis, Acute Necrotizing/complications*
;
Acute Disease
;
Intraabdominal Infections/complications*
;
Necrosis/complications*
;
Treatment Outcome

Result Analysis
Print
Save
E-mail