Introduction: Acute liver injury (ALI) is a common complication of sepsis and is associated with adverse clinical outcomes. We aimed to develop a model to predict the risk of ALI in patients with sepsis after hospitalization. Methods: Medical records of 3196 septic patients treated at the Lishui Central Hospital in Zhejiang Province from January 2015 to May 2023 were selected. Cohort 1 was divided into ALI and non-ALI groups for model training and internal validation. The initial laboratory test results of the study subjects were used as features for machine learning (ML), and models built using nine different ML algorithms were compared to select the best algorithm and model. The predictive performance of model stacking methods was then explored. The best model was externally validated in Cohort 2. Results: In Cohort 1, LightGBM demonstrated good stability and predictive performance with an area under the curve (AUC) of 0.841. The top five most important variables in the model were diabetes, congestive heart failure, prothrombin time, heart rate, and platelet count. The LightGBM model showed stable and good ALI risk prediction ability in the external validation of Cohort 2 with an AUC of 0.815. Furthermore, an online prediction website was developed to assist healthcare professionals in applying this model more effectively. Conclusions: The Light GBM model can predict the risk of ALI in patients with sepsis after hospitalization.

BackgroundAcute kidney injury (AKI) is not only a complication but also a serious threat to patients with cerebral infarction (CI). This study aimed to explore the application of interpretable machine learning algorithms in predicting AKI in patients with cerebral infarction.MethodsThe study included 3,920 patients with CI admitted to the Intensive Care Unit and Emergency Medicine of the Central Hospital of Lishui City, Zhejiang Province. Nine machine learning techniques, including XGBoost, logistics, LightGBM, random forest (RF), AdaBoost, GaussianNB (GNB), Multi-Layer Perceptron (MLP), support vector machine (SVM), and k-nearest neighbors (KNN) classification, were used to develop a predictive model for AKI in these patients. SHapley Additive exPlanations (SHAP) analysis provided visual explanations for each patient. Finally, model effectiveness was assessed using metrics such as average precision (AP), sensitivity, specificity, accuracy, F1 score, precision-recall (PR) curve, calibration plot, and decision curve analysis (DCA).ResultsThe XGBoost model performed better in the internal validation set and the external validation set, with an AUC of 0.940 and 0.887, respectively. The five most important variables in the model were, in order, glomerular filtration rate, low-density lipoprotein, total cholesterol, hemiplegia and serum kalium.ConclusionThis study demonstrates the potential of interpretable machine learning algorithms in predicting CI patients with AKI.

Objective:To investigate the effect of sodium butyrate (NaB) on renal and intestinal injury after cardiac arrest and cardiopulmonary resuscitation (CA-CPR) and its related mechanism.Methods:Twenty-four domestic healthy male swines were randomly divided into 3 groups: sham group ( n=6), CA-CPR group ( n=10) and NaB group ( n=8). The animals only underwent operational preparation in the sham group. The animal model of CA and CPR was established by 9 min of ventricular fibrillation induced by electrical stimulation in the ventricle and then 6 min of CPR in the CA-CPR and NaB groups. At 5 min after resuscitation, a dose of 75 mg/kg of NaB was intravenously infused for 1 h in the NaB group, and meanwhile the same volume of vehicle was intravenously infused in the sham and CA-CPR groups. At 1, 2, 4, and 24 h after resuscitation, blood samples were collected to detect the renal and intestinal injury biomarkers, such as creatinine (Cr), blood urea nitrogen (BUN), intestinal fatty acid binding protein (IFABP), and diamine oxidase (DAO). At 24 h after resuscitation, renal and intestinal tissue specimens were harvested to detect the protein markers of cell autophagy including microtubule-associated protein light chain 3 Ⅱ (LC3Ⅱ) and p62 expression, and also renal and intestinal apoptosis. Statistical analysis was performed by SPSS software, and continuous variables were compared with one-way analysis of variance among the groups. Results:After CA-CPR, the renal and intestinal injury biomarkers including Cr, BUN, IFABP, and DAO were significantly increased at all time points after resuscitation in the CA-CPR and NaB groups compared with the sham group (all P<0.05). The injury biomarkers mentioned-above were significantly lower at all time points after resuscitation in the NaB group than in the CA-CPR group [Cr (μmol/L): (90±5) vs. (127±9) at 1 h, (135±14) vs. (168±9) at 2 h, (174±10) vs. (211±12) at 4 h, (192±10) vs. (253±13) at 24 h; BUN (mmol/L): (10.5±1.0) vs. (12.3±1.0) at 1 h, (12.2±1.2) vs. (15.3±0.9) at 2 h, (13.6±1.3) vs. (18.3±1.2) at 4 h, (15.4±1.4) vs. (21.5±1.4) at 24 h; IFABP (pg/mL): (502±33) vs. (554±32) at 1 h, (574±52) vs. (644±41) at 2 h, (646±44) vs. (732±43) at 4 h, (711±42) vs. (828±42) at 24 h; DAO (U/mL): (8.6±1.0) vs. (10.5±0.9) at 1 h, (10.6±1.2) vs. (12.8±1.0) at 2 h, (12.1±1.0) vs. (15.0±1.0) at 4 h, (14.1±1.1) vs. (17.6±1.0) at 24 h, (all P<0.05)]. Renal and intestinal tissue detection indicated that cell autophagy and apoptosis were significantly increased after resuscitation in the CA-CPR and NaB groups compared with the sham group, which was indicated by significantly increased LC3Ⅱ and decreased p62 expression, and markedly elevated apoptosis index (all P<0.05). However, cell autophagy and apoptosis in the kidney and intestine were significantly milder after resuscitation in the NaB group than in the CA-CPR group [renal LC3 Ⅱ: (1.15±0.17) vs. (2.23±0.31), p62: (1.60±0.10) vs. (1.17±0.08), apoptosis index (%): (21.2±5.3) vs. (50.9±7.9); intestinal LC3 Ⅱ: (1.03±0.17) vs. (1.71±0.21), p62: (1.30±0.29) vs. (0.79±0.29), apoptosis index (%): (25.6±6.1) vs. (61.7±10.7), all P<0.05]. Conclusions:NaB could alleviate the severity of renal and intestinal damage after CA-CPR in swine, and its protective mechanism may be related to the inhibition of cell autophagy and apoptosis.

Objective:To explore the protective effect and mechanism of sodium butyrate on cardio-brain injury after cardiopulmonary resuscitation in pigs.Methods:A total of 24 large white healthy male pigs were divided into a sham operation group (n = 6), a cardiopulmonary resuscitation group (n = 10) and a sodium butyrate group (n = 8) by a random number method. In the sham operation group, only tracheal intubation and arteriovenous catheterization were performed, and induced ventricular fibrillation and cardiopulmonary resuscitation were not performed. Cardiac arrest and cardiopulmonary resuscitation models were prepared by electrical stimulation for 9 min and cardiopulmonary resuscitation for 6 min. At 5 min after successful resuscitation, the sodium butyrate group was injected 75 mg/kg of sodium butyrate through the femoral vein for 1 h. Sham operation and cardiopulmonary resuscitation groups were given the same amount of isotonic NaCl solution by the same method. Serum levels of cardiac troponin I (cTnI), creatine kinase MB (CK-MB), neuron specific enolase (NSE) and S100 calcium-binding protein B (S100B) were measured by the enzyme-linked immunosorbent assay (ELISA) before cardiac arrest and at 1, 2, 4, and 24 h after resuscitation. At 24 h after resuscitation, the expression levels of microtubule-associated protein 1 light chain 3 (LC3) and ubiquitin-binding protein 62 (p62) in the cardiac muscle and cerebral cortex tissue were detected by western-blotting, and the apoptosis rate was detected by the TUNEL method.Results:In the cardiopulmonary resuscitation group, nine pigs were successfully resuscitated, seven of which survived to 24 h. In the sodium butyrate group, seven pigs were resuscitated successfully, and all survived to 24 h. There was no significant difference in the success of 24-h resuscitation among the three groups (χ2 = 3.381, P = 0.184). At 1, 2, 4 and 24 h after resuscitation, serum cTnI (F = 62.736, 201.265, 330.083, 643.538; all P < 0.001), CK-MB (F = 90.380, 181.548, 669.657, 707.485; all P < 0.001), NSE (F = 92.302, 125.730, 590.627, 1 130.372; all P < 0.001) and S100B (F = 42.831, 152.004, 295.517, 1 023.409; all P < 0.001) were observed and compared in the three groups, and the differences were statistically significant. Compared with the sham operation group, the cTnI, CK-MB, NSE and S100B levels in serum were significantly increased at 2, 4 and 24 h after resuscitation in the cardiopulmonary resuscitation group and sodium butyrate group, which were highest in the cardiopulmonary resuscitation group (all P < 0.05). At 24 h after resuscitation, the levels of autophagy markers LC3 (F = 81.671, 49.204; both P < 0.001) and p62 (F = 127.620, 65.594; both P < 0.001) and the apoptosis rate (F = 116.750, 249.105; both P < 0.001) in cardiomyocytes and cerebral cortex cells were observed and compared in the three groups, and the differences were statistically significant. Compared with the sham operation group, the LC3 expression and apoptosis index were increased, and the p62 expression was decreased in myocardium and cerebral cortex of pigs in the cardiopulmonary resuscitation group and sodium butyrate group at 24 h after resuscitation (all P < 0.05). However, compared with the cardiopulmonary resuscitation group, the LC3 expression and apoptosis index were decreased, and the p62 expression was increased in the myocardium and cerebral cortex in the sodium butyrate group at 24 h after resuscitation (all P < 0.05).Conclusion:Sodium butyrate has a protective effect on cardio-brain injury after resuscitation of cardiac arrest, and the mechanism may be related to inhibition of autophagy and apoptosis.

Objective To explore the protective effects of sulforaphane（SFN） on acute lung injury（ALI） after cardiopulmonary resuscitation in pigs and its potential mechanism. Methods Twenty-four conventional white pigs were chosen and were randomly divided into the following groups by using a random number table: sham group（n=6）, model group（n=9）, and SFN group（n=9）. The study chose to establish the porcine model of ventricular fibrillation-related cardiac arrest（CA） and cardiopulmonary resuscitation（CPR） based on the methods including 10 min of CA electrically induced by the right ventricle and then 6 min of CPR. At 5 min after the animal model establishment, a dose of 2 mg/kg SFN was infused via the femoral vein within 10 min in the SFN group. Prior to the model establishment and during 4 h observation after the model establishment, the changes of extravascular lung water index（ELWI） and pulmonary vascular permeability index（PVPI） were regularly measured by using PiCCO monitor. At 24 h after the model establishment, the animals were sacrificed and their left lower lung lobe’s tissue specimens were obtained to detect the index of cell apoptosis by TUNEL assay, and the protein expression levels of cleaved caspase 3, receptor-interacting protein 3（RIP3）, and mixed lineage kinase domain-like protein（MLKL） by Western blot. Results（1）Prior to the model establishment, the values of ELWI and PVPI were kept at an even physiological level in the three groups（all P＞0.05）. After the model establishment, the values of ELWI and PVPI at different time points in the model group were significantly increased, and the values of ELWI within 2 h and the values of PVPI at different time points in the SFN group were significantly elevated when compared with the sham group（all P＜0.05）. However, the values of ELWI at 2 h after the model establishment and the values of PVPI at 1 h after the model establishment were significantly decreased in the SFN group when compared to the model group（all P＜0.05）.（2） At 24 h after the model establishment, the index of cell apoptosis was significantly increased, and the protein expression levels of cleaved caspase 3, RIP3, and MLKL were significantly elevated in the lung tissues in the model and SFN groups when compared with the sham group（all P＜0.05）. However, the index of cell apoptosis was significantly decreased, and the protein expression levels of cleaved caspase 3, RIP3, and MLKL were significantly reduced in the lung tissues in the SFN group when compared to the model group（all P＜0.05）. Conclusions SFN could effectively alleviate the severity of ALI after CPR in pigs and produce the effective lung protection, in which the protective approach is related to the inhibition of cell apoptosis and necroptosis in lung tissues.

目的 通过分析心肺复苏患者在高级生命支持阶段的机械通气时气道峰压水平,为选择合理的气道高压报警上限提供依据.方法 选择进行心肺复苏时间>20min的患者,使用伟康Eisprit呼吸机,合理设定呼吸机其他工作参数,初始设置气道高压报警值为40cmH2O,如果患者气道峰压>40cmH2O则上调至50cmH2O,如果>50cmH2O则上调至60cmH2O为止不再上调,用呼吸机自带的测量软件测量患者的气道峰压高值,并记录气胸等不良反应的发生情况,共完成60例患者的数据收集和分析.结果 气道峰压>40cmH2O的患者占76.7％,<50cmH2O患者占81.6％,56～60cmH2O患者占5％,6.7％患者>60cmH2O.有2例气道峰压>55cmH2O的患者在12h内出现少量气胸.结论 心肺复苏高级生命支持期间的机械通气,因患者气道峰压普遍高于常规设置气道高压报警而容易发生人机不协调,据本次资料,调整呼吸机参数将患者气道峰压控制在<55cmH2O,以避免气压伤的发生是有必要的.而将气道高压报警设置为50cmH2O,既能保证患者的气道安全,又不会引起过多不必要的报警事件,也是比较合适的初始设置.

目的 观察介入诊疗术对创伤所致腹部急性大出血的临床应用价值.方法 采用改良Seldinger技术经股动脉穿刺插管对91例创伤所致腹腔急性大出血患者进行选择性血管造影,并进行栓塞治疗.结果 经选择性血管造影,91例急性大出血患者均发现出血征象,诊断阳性率为100％.91例栓塞治疗后88例止血效果良好,总效率为96.7％.2例脾破裂患者术后第2天出现再次出血,其中1例出现失血性休克表现,转手术室行开腹脾切除术,另1例生命体征稳定,出血量不大,给予再次栓塞术后出血停止.91例患者进行3～36个月的随访,平均随访时间18个月,未发现后遗症.结论 集血管造影诊断和栓塞治疗于一体的介入诊疗技术已成为治疗急性大出血的首要治疗方法之一,具有微创、出血灶定位准、止血迅速及复发率低等诸多优势,对急诊抢救患者生命发挥了积极重要的作用.