A novel interpretative tool for early prediction of low cardiac output syndrome after valve surgery: online machine learning models.

OBJECTIVE: Low cardiac output syndrome (LCOS) is a severe complication after valve surgery, with no uniform standard for early identification. We developed interpretative machine learning (ML) models for predicting LCOS risk preoperatively and 0.5 h postoperatively for intervention in advance. METHODS: A total of 2218 patients undergoing valve surgery from June 2019 to Dec 2021 were finally enrolled to construct preoperative and postoperative models. Logistic regression, support vector machine (SVM), random forest classifier, extreme gradient boosting, and deep neural network were executed for model construction, and the performance of models was evaluated by area under the curve (AUC) of the receiver operating characteristic and calibration curves. Our models were interpreted through SHapley Additive exPlanations, and presented as an online tool to improve clinical operability. RESULTS: The SVM algorithm was chosen for modeling due to better AUC and calibration capability. The AUCs of the preoperative and postoperative models were 0.786 (95% CI 0.729-0.843) and 0.863 (95% CI 0.824-0.902), and the Brier scores were 0.123 and 0.107. Our models have higher timeliness and interpretability, and wider coverage than the vasoactive-inotropic score, and the AUC of the postoperative model was significantly higher. Our preoperative and postoperative models are available online at http://njfh-yxb.com.cn:2022/lcos. CONCLUSIONS: The first interpretable ML tool with two prediction periods for online early prediction of LCOS risk after valve surgery was successfully built in this study, in which the SVM model has the best performance, reserving enough time for early precise intervention in critical care.

Association of Critical Value With 28-Day Mortality After Cardiac Surgery.

OBJECTIVE: The emergence of critical values gives a warning to the medical safety of hospitalized patients, especially Cardiosurgery Intensive Care Unit (CSICU) patients. The aim of this study was to investigate the association between early postoperative critical values and the prognosis of patients after cardiac surgery. METHODS: Clinical data of the patients were obtained from the Cardiac Critical Care Clinical Database of the Cardiovascular Intensive Care Unit of Nanjing First Hospital. A total of 1,598 consecutive patients undergoing cardiac surgery were enrolled in this retrospective cohort study, during the period from July 2019 to December 2020. According to whether critical value occurred within 7 days after cardiac surgery, patients were divided into two groups: the critical value group and control group. COX regression and survival analysis were performed to analyze the clinical data of the two groups. The area under the receiver operating characteristic curve (ROC) was used to assess the critical value's predictive value and determine the optimal cutoff value. RESULTS: With patients in the critical value group, the 28-day mortality after cardiac surgery was 21.98%, significantly higher than that of the control group (P < 0.05). Logistic regression analysis revealed the APACHE II score (Adjusted HR-1.11, 95% CI-1.043-1.185) and critical value group (Adjusted HR-13.57, 95% CI-6.714-27.435 ) were independent predictors of 28-day mortality after cardiac surgery. The ROC curve showed that the critical value case model (AUC = 0.748 ± 0.052, P < 0.05) could effectively predict the 28-day mortality, and the optimum cutoff was 1 case (sensitivity 52.63%, specificity 95.70%). CONCLUSIONS: One or more reported cases of critical values in the early postoperative period could be an independent risk factor for 28-day mortality in patients undergoing cardiac surgery. The predictive model based on critical value might be effective in clinical therapy and risk stratification.

Association between Left Ventricular Longitudinal Strain (GLS) and Prognosis of the Patients Undergoing Heart Valve Surgery with Preserved Left Ventricular Ejection Fraction.

PURPOSE: Global longitudinal strain (GLS) seems accurate for detecting subclinical myocardial dysfunction. This study aimed to determine the association between GLS and postoperative intensity of inotropic support in the patients undergoing heart valve surgery with preserved left ventricular ejection fraction. METHODS: 74 patients with preserved left ventricular ejection fraction who underwent valve surgery during the period between March 2021 and June 2022 were included in this prospective observational study. Transthoracic echocardiography including strain analysis with speckle tracking was performed before surgery. Patients were stratified according to the left ventricle (LV) GLS: LV-GLS ≥-16% (Impaired GLS group) and LV-GLS <-16% (Normal GLS group). The primary endpoint was postoperative vasoactive inotropic score. A high vasoactive inotropic score (VIS) was defined as a maximum VIS of ≥15 within 24 hours postoperatively. Postoperative adverse events, baseline clinical and echocardiographic data were also recorded. We invested the ability of preoperative GLS in predicting adverse postoperative outcomes, such as prolonged mechanical ventilation and the need for pharmacologic hemodynamic support after cardiac surgery. RESULTS: Seventy-four patients were included and analyzed in this study, including thirty-three in impaired GLS group and forty-one in normal GLS group. In-hospital mortality was 1.27% (1/74). Patients in impaired GLS group were more likely to have prolonged mechanical ventilation (p = 0.041). Multivariable logistic regression analysis revealed that the apical four-chamber view of the left ventricle (A4C)-GLS was significantly associated with high VIS (OR 1.373, p = 0.007). A4C-GLS had a sensitivity of 62.5% and a specificity of 89.66% for predicting high VIS (area under the curve, 0.78). The relationships between GLS and other secondary outcome measures were not statistically significant. The optimal cutoff of A4C-GLS for postoperative high vasoactive inotropic score was -10.85%. CONCLUSION: Preoperative LV dysfunction is an independent risk factor for postoperative high VIS. A4C-GLS may be a reliable tool in predicting high VIS after cardiac surgery. Those patients with impaired contractility were at high risk for elevated inotropic support and prolonged mechanical ventilation after cardiac surgery. These findings suggest an important role for echocardiographic GLS in perioperative assessment of cardiac function in the patients undergoing cardiac surgery.

Effect of PA-MSAH preprocessing on the expression of TLR-4-NF-κB pathway and inflammatory factors in the intestinal tract of rats with septic shock.

Effects of Pseudomona blank s aeruginosa-mannose sensitive hemagglutini (PA-MSHA) preprocessing on toll like receptor (TLR)-4-NF-κB pathway and inflammatory factors expression in the intestinal tract of rats with septic shock were investigated. A total of 30 rats were randomly divided into 3 groups (n=10): Blank control, septic shock, and PA-septic shock group. After the model was successfully established, the average arterial pressure in rats was monitored. The concentration of cytokine interleukin-l (IL-1), IL-6 and tumor necrosis factor-α (TNF-α) were determined by ELISA method. Flow cytometry was performed to detect TLR-4 expression. Number of in vitro chemotaxised neutrophils was detected by Transwell chamber. The expression of TLR-4, NF-κB and ICAM-1 and VCAM-1 was detected by western blot analysis. The concentration of cytokine IL-1, IL-6, TNF-α in the peritoneal lavage fluid and the intestinal tissue significantly increased in the septic shock group and the PA-septic shock group (P<0.05), and the concentration in the PA-septic shock group was significantly lower than that of the septic shock group (P<0.05). Compared to the control group, the expression of TLR-4, NF-κB and ICAM-1 and VCAM-1 increased in the septic shock and PA-septic shock group (P<0.05), and the expression level of PA-septic shock group was lower than the septic shock group (P<0.05). The expression of TLR-4 in the PA-septic shock group was lower than the septic shock group (P<0.05). PA-MSHA pretreatment reduced inflammation, thus preventing the intestinal injury caused by septic shock.

Percutaneous catheter drainage combined with peritoneal dialysis for treating acute severe pancreatitis: a single-center prospective study.

BACKGROUND: To investigate the efficacy of percutaneous catheter drainage (PCD) and peritoneal dialysis (PD) in the treatment of severe acute pancreatitis (SAP) and its underlying mechanism. METHODS: Totally 64 SAP patients were included in our study and randomly assigned into PCD+PD group (the combination group, N.=32) and convention group (N.=32). SAP patients in the combination group were treated with percutaneous catheter drainage combined with peritoneal dialysis, while those in the convention group were treated with conventional method. The treatment efficacy of both methods were evaluated by comparing levels of plasma inflammatory cytokines (IL-6, IL-8, TNF-α, C-reactive protein, procalcitonin and leukocyte count), relative indexes of important organs (aspartate aminotransferase, alanine aminotransferase, creatinine and urea nitrogen) and other clinical data (amelioration time of abdominal pain and abdominal distension, Balthazar CT scores, acute physiology and chronic health enquiry II score, length of hospital stay, complications and prognosis). RESULTS: The expression levels of inflammatory cytokines were significantly decreased in the combination group in a time-dependent manner in comparison with those of the convention group. In addition, the amelioration time of abdominal pain and abdominal distension, length of hospital stay, Balthazar CT scores and the acute physiology and chronic health care II scores in the combination group were also significantly decreased in comparison with those of the convention group. CONCLUSIONS: The combination treatment of PCD and PD effectively relieves the clinical symptoms of SAP by clearing plasma inflammatory cytokines.

Early diagnostic value of urinary NGAL for postoperative acute kidney injury in patients with acute Stanford type A aortic dissection / 中国胸心血管外科临床杂志

@#Objective     To investigate the early diagnostic value of urinary neutrophil gelatinase-associated lipocalin (NGAL) for acute kidney injury (AKI) after acute Stanford type A aortic dissection. Methods     From January 2018 to December 2018, the clinical data of 50 patients who underwent open surgery for acute Stanford type A aortic dissection were analyzed in Nanjing First Hospital. Urine specimens were collected before and 2 hours after the aortic dissection surgery. Patients were divided into an AKI group (n=27) and a non-AKI group (n=23) according to the Kidney Disease Improving Global Outcomes criteria. Receiver operating characteristic (ROC) curve was used to evaluate the diagnostic value of urine NGAL.  Results    The incidence of postoperative AKI was 54.00% (27/50). There was a statistically significant difference between the two groups in serum creatinine concentration at 2 hours after surgery and urinary NGAL concentration before the surgery (P<0.05). The area under ROC curve of preoperative urinary NGAL concentration was 0.626. When cut-off value was 43 ng/mL, the sensitivity was 40.7%, specificity was 95.7%. The area under ROC curve of urinary NGAL concentration at 2 hours after surgery was 0.655, and when the cut-off value was 46.95 ng/mL, the sensitivity was 63.0%, specificity was 78.3%.  Conclusion     Urine NGAL can predict postoperative AKI in patients with acute Stanford type A aortic dissection, but its value is limited.