OBJECTIVE: To identify the optimal machine learning algorithm for predicting post-stroke epilepsy (PSE) within one year following acute stroke, establish a nomogram model based on this algorithm, and perform external validation to achieve accurate prediction of secondary epilepsy. METHODS: A total of 870 acute stroke patients admitted to the emergency department of Xiang'an Hospital of Xiamen University from June 2019 to June 2023 were enrolled for model development (model group). An external validation cohort of 435 acute stroke patients admitted to the Fifth Hospital of Xiamen during the same period was used to validate the machine learning algorithms and nomogram model. Patients were classified into control and epilepsy groups based on the development of PSE within one year. Clinical and laboratory data, including baseline characteristics, stroke location, vascular status, complications, hematologic parameters, and National Institutes of Health Stroke Scale (NIHSS) score, were collected for analysis. Nine machine learning algorithms such as logistic regression, CN2 rule induction, K-nearest neighbors, adaptive boosting, random forest, gradient boosting, support vector machine, naive Bayes, and neural network were applied to evaluate predictive performance. The area under the curve (AUC) of receiver operator characteristic curve (ROC curve) was used to identify the optimal algorithm. Logistic regression was used to screen risk factors for PSE, and the top 10 predictors were selected to construct the nomogram model. The predictive performance of the model was evaluated using the ROC curve in both the model and validation groups. RESULTS: Among the 870 patients in the model group, 29 developed PSE within one year. Among the nine algorithms tested, logistic regression demonstrated the best performance and generalizability, with an AUC of 0.923. Univariate logistic regression identified several risk factors for PSE, including platelet count, white blood cell count, red blood cell count, glycated hemoglobin (HbA1c), C-reactive protein (CRP), triglycerides, high-density lipoprotein (HDL), aspartate aminotransferase (AST), alanine aminotransferase (ALT), activated partial thromboplastin time (APTT), thrombin time, D-dimer, fibrinogen, creatine kinase (CK), creatine kinase-MB (CK-MB), lactate dehydrogenase (LDH), serum sodium, lactic acid, anion gap, NIHSS score, brain herniation, periventricular stroke, and carotid artery plaque. Further multivariate logistic regression analysis showed that white blood cell count, HDL, fibrinogen, lactic acid and brain herniation were independent risk factors [odds ratio (OR) were 1.837, 198.039, 47.025, 11.559, 70.722, respectively, all P < 0.05]. In the external validation group, univariate logistic regression analysis showed that platelet count, white blood cell count, CRP, triacylglycerol, APTT, D-dimer, fibrinogen, CK, CK-MB, LDH, NIHSS score, and cerebral herniation were risk factors for PSE one year after acute stroke. Further multiple logistic regression analysis showed that APTT and cerebral herniation were independent predictors (OR were 0.587 and 116.193, respectively, both P < 0.05). The nomogram model, constructed using 10 key variables-brain herniation, periventricular stroke, carotid artery plaque, white blood cell count, triglycerides, thrombin time, D-dimer, serum sodium, lactic acid, and NIHSS score-achieved an AUC of 0.908 in the model group and 0.864 in the external validation group. CONCLUSIONS The logistic regression-based prediction model for epilepsy one year after acute stroke, developed using machine learning algorithms, showed optimal predictive performance. The nomogram model based on the logistic regression-derived predictors showed strong discriminative power and was successfully validated externally, suggesting favorable clinical applicability and generalizability.
Humans ; Machine Learning ; Stroke/complications* ; Nomograms ; Epilepsy/etiology* ; Algorithms ; Male ; Female ; Logistic Models ; Middle Aged ; Aged ; Risk Factors ; Bayes Theorem

OBJECTIVE: To develop and compare risk prediction models for in-hospital post-cardiac arrest brain injury (PCABI) in critically ill patients using nomograms and random forest algorithms, aiming to identify the optimal model for early identification of high-risk PCABI patients and providing evidence for precise treatment. METHODS: A retrospective cohort study was used to collect the first-time in-hospital cardiac arrest (IHCA) patients admitted to the intensive care unit (ICU) from 2008 to 2019 in the Medical Information Mart for Intensive Care-IV (MIMIC-IV) as the study population, and the patients' age, gender, body mass, health insurance utilization, first vital signs and laboratory tests within 24 hours of ICU admission, mechanical ventilation, and critical care scores were extracted. Independent influencing factors of PCABI were identified through univariate and multivariate Logistic regression analyses. The included patients were randomly divided into a training cohort and an internal validation cohort in a 7:3 ratio, and the PCABI risk prediction model was constructed by the nomogram and random forest algorithm, respectively, and the model was evaluated by receiver operator characteristic curve (ROC curve), the calibration curve, and the decision curve analysis (DCA), and after the better model was selected, 179 patients admitted to Tianjin Medical University General Hospital as the external validation cohort for external evaluation were collected by using the same inclusion and exclusion criteria. RESULTS: A total of 1 419 patients with without traumatic brain injury who had their first-time IHCA were enrolled, including 995 in the training cohort (including 176 PCABI and 819 non-PCABI) and 424 in the internal validation cohort (including 74 PCABI and 350 non-PCABI). Univariate and multivariate analysis showed that age, potassium, urea nitrogen, sequential organ failure assessment (SOFA), acute physiology and chronic health evaluation III (APACHE III), and mechanical ventilation were independent influences on the occurrence of PCABI in patients with IHCA (all P < 0.05). Combining the above variables, we constructed a nomogram model and a random forest model for comparison, and the results show that the nomogram model has better predictive efficacy than the random forest model [nomogram model: area under the ROC curve (AUC) of the training cohort = 0.776, with a 95% credible interval (95%CI) of 0.741-0.811; internal validation cohort AUC = 0.776, with a 95%CI of 0.718-0.833; random forest model: AUC = 0.720, with a 95%CI of 0.653-0.787], and they performed similarly in terms of calibration curves, but the nomogram performed better in terms of decision curve analysis (DCA); at the same time, the nomogram model was robust in terms of external validation cohort (external validation cohort AUC = 0.784, 95%CI was 0.692-0.876). CONCLUSIONS A nomogram risk prediction model for the occurrence of PCABI in critically ill patients was successfully constructed, which performs better than the random forest model, helps clinicians to identify the risk of PCABI in critically ill patients at an early stage and provides a theoretical basis for early intervention.
Humans ; Critical Illness ; Retrospective Studies ; Heart Arrest/complications* ; Nomograms ; Brain Injuries/etiology* ; Intensive Care Units ; Algorithms ; Male ; Female ; Middle Aged ; ROC Curve ; Risk Factors ; Risk Assessment ; Logistic Models ; Aged

OBJECTIVE: To develop and validate a nomogram model for predicting 30-day mortality among elderly patients with sepsis-associated liver dysfunction (SALD), to identify high-risk patients and improve prognosis. METHODS: A retrospective cohort study was conducted using data extracted from the Medical Information Mart for Intensive Care-IV (MIMIC-IV) database for elderly patients with SALD who were first admitted to the intensive care unit (ICU) of Beth Israel Deaconess Medical Center between 2008 and 2019, including basic characteristics, severity scores, underlying diseases, infection foci, 24-hour vital signs, initial laboratory indicators, 24-hour complications, and prognosis related indicators. Patients were randomly assigned to training group and validation group in a ratio of 7 : 3. The training group used the LASSO regression analysis, as well as multivariate Logistic regression analysis to screen for independent risk factors for 30-day mortality. A nomogram prediction model was constructed, and receiver operator characteristic curve (ROC curve), calibration curves, and decision curve analysis (DCA) were used to evaluate the model, and validate the model using the validation cohort. RESULTS: A total of 630 elderly patients with SLAD were included in the study, including 441 in the training group and 189 in the validation group. Oxford acute severity of illness score (OASIS) for training group [odds ratio (OR) = 1.060, 95% confidence interval (95%CI) was 1.034-1.086], 24-hour pulse oxygen saturation (SpO₂; OR = 0.876, 95%CI was 0.797-0.962), initial mean corpuscular volume (MCV; OR = 1.043, 95%CI was 1.009-1.077), initial red blood cell distribution width (RDW; OR = 1.237, 95%CI was 1.123-1.362), initial blood glucose (OR = 1.008, 95%CI was 1.004-1.013), and initial aspartate aminotransferase (AST; OR = 1.000, 95%CI was 1.000-1.001) were independent risk factors for 30-day mortality in patients (all P < 0.05). Based on the above variables, a nomogram model was constructed, and the ROC curve showed that the area under the curve (AUC) of the model in the training group was 0.757 (95%CI was 0.712-0.803), with a sensitivity of 65.05% and a specificity of 74.90%; the AUC of the model in the validation group was 0.712 (95%CI was 0.631-0.792), with a sensitivity of 58.67% and a specificity of 81.58%. The calibration curves of the training and validation groups show that both the fitted curves were close to the standard curves. The Hosmer-Lemeshow test: the training group (χ ² = 6.729, P = 0.566), the validation group (χ ² = 13.889, P = 0.085), indicating that the model can fit the observed data well. The DCA curve shows that when the threshold probability of the training group was 16% to 94% and the threshold probability of the validation group was 27% to 99%, the net benefit of the model was good. CONCLUSIONS OASIS, 24-hour SpO₂, initial MCV, initial RDW, initial blood glucose and initial AST are independent risk factors for 30-day mortality in elderly patients with SALD. The nomogram based on these six variables demonstrates good predictive performance.
Humans ; Sepsis/complications* ; Retrospective Studies ; Nomograms ; Aged ; Prognosis ; Risk Factors ; Liver Diseases/mortality* ; Intensive Care Units ; ROC Curve ; Male ; Female ; Logistic Models

OBJECTIVE: To construct and validate a prognostic prediction model for children with sepsis using the Phoenix sepsis score (PSS). METHODS: A retrospective case series study was conducted to collect clinical data of children with sepsis admitted to the pediatric intensive care unit (PICU) of the Affiliated Hospital of Guizhou Medical University from January 2022 to April 2024. The data included general information, the worst values of laboratory indicators within the first 24 hours of PICU admission, PSS score, pediatric critical illness score (PCIS), and the survival status of the children within 30 days of admission. The statistically significant indicators in univariate Logistic regression analysis were included in multivariate Logistic regression analysis to screen the risk factors affecting the prognosis of children with sepsis and construct a nomogram model. The receiver operator characteristic curve (ROC curve) was drawn to evaluate the predictive performance of the model. The Bootstrap method was used to perform 1 000 repeated sampling internal verification and draw the calibration curve of the model. RESULTS: A total of 199 children with sepsis were included, of which 32 died and 167 survived 30 days after admission. In the univariate Logistic regression analysis, shock, white blood cell count (WBC), international normalized ratio (INR), lactic acid (Lac), PSS score, and PCIS score were identified as statistically significant predictors. These variables were then included in the multivariate Logistic regression analysis, which demonstrated that shock [odds ratio (OR) = 4.258, 95% confidence interval (95%CI) was 1.049-17.288], WBC (OR = 1.124, 95%CI was 1.052-1.210), and PSS score (OR = 1.977, 95%CI was 1.298-3.012) were independent risk factors for mortality in pediatric patients with sepsis (all P < 0.05). A nomogram model was constructed based on these three risk factors, with the model equation as follows: -4.809+1.449×shock+0.682×PSS score+0.117×WBC. The calibration curve results showed that the model's predictions were highly consistent with the actual observations. The ROC curve showed that when the Youden index of the prediction model was 0.792, the sensitivity and specificity were 90.6% and 88.6%, respectively, and the area under the curve (AUC) was 0.957 (95%CI was 0.930-0.984), which was higher than the AUC of shock, WBC, and PSS score alone (0.808, 0.667, 0.908, respectively). CONCLUSIONS Shock, WBC, and PSS score have demonstrated certain predictive value for mortality in children with sepsis. The nomogram model based on the above indicators has important clinical significance for evaluating the prognosis and guiding treatment of children with sepsis.
Humans ; Sepsis/diagnosis* ; Prognosis ; Retrospective Studies ; Logistic Models ; Intensive Care Units, Pediatric ; Nomograms ; Child ; ROC Curve ; Risk Factors ; Male ; Female ; Child, Preschool ; Infant

OBJECTIVE: To explore the timing of weaning extracorporeal membrane oxygenation (ECMO) and analyze the risk factors that affect survival outcomes before weaning. METHODS: A retrospective case-control study was conducted. Patients who received ECMO treatment and were weaned according to physicians' orders at the Second Xiangya Hospital of Central South University from January 2020 to June 2024 were enrolled as the study subjects. The general information, underlying diseases, indications and processes of ECMO, vital signs and arterial blood gas analysis 1 hour before weaning test, and biochemical indicators 24 hours before weaning test were collected through the hospital electronic medical record system. The primary outcome measure was the hospital mortality. The variables with P < 0.1 in univariate analysis and correlation analysis were included into binary Logistic regression analysis to identify risk factors. A nomogram model was constructed to predict the risk of weaning death in patients with ECMO, and receiver operator characteristic curve (ROC curve) and calibration curve were drawn to evaluate the model. Decision curve analysis (DCA) was used to evaluate the clinical net benefit rate of the model. RESULTS: A total of 32 ECMO patients were included, among whom 10 received veno-arterial ECMO (VA-ECMO) and 22 received veno-venous ECMO (VV-ECMO). During the hospitalization period, 23 patients survived, while 9 died. The time from mechanical ventilation to ECMO activation in the death group was significantly longer than that in the survival group, and the time from ECMO cessation to discharge was significantly shorter than that in the survival group. The levels of diastolic blood pressure (DBP) and albumin (Alb) before weaning were significantly lower than those in the survival group, and the level of procalcitonin (PCT) was significantly higher than that in the survival group (all P < 0.05). Spearman correlation analysis showed that DBP, PCT, Alb, and thrombin time (TT) were correlated with the weaning outcomes of ECMO patients (r values were -0.450, 0.373, -0.376, -0.346, all P < 0.1). Binary Logistic regression analysis showed that the final indicators entering the regression equation included DBP [odds ratio (OR) = 0.864, 95% confidence interval (95%CI) was 0.756-0.982], PCT (OR = 1.157, 95%CI was 0.679-1.973), and TT (OR = 0.852, 95%CI was 0.693-1.049), and a nomogram model was constructed to predict the weaning outcomes of ECMO patients. ROC curve analysis showed that the area under the curve (AUC) of the nomogram model for predicting the weaning outcome of ECMO patients was 0.831, with a sensitivity of 77.8% and a specificity of 65.2%. Its predictive value was better than that of single indicators DBP, PCT, and TT (AUC of 0.787, 0.739, and 0.722, respectively). The calibration curve showed that the prediction probability of the model was in good consistency with the actual observed results, the Hosmer-Lemeshow goodness of fit test showed that, χ ² = 8.3521, P = 0.400, indicating that the model fits well. DCA showed that across risk threshold of 0-0.8, the net benefit rate was greater than 0, which was significantly better than that of single indicator. CONCLUSIONS The nomogram model constructed with DBP, PCT, and TT has certain predictive value for the weaning outcomes of ECMO patients and can be used as a screening indicator for ECMO weaning timing.
Humans ; Extracorporeal Membrane Oxygenation ; Retrospective Studies ; Risk Factors ; Case-Control Studies ; Hospital Mortality ; Male ; Female ; Nomograms ; Logistic Models ; ROC Curve ; Middle Aged ; Adult ; Ventilator Weaning ; Time Factors

OBJECTIVE: To identify the risk factors for acute respiratory distress syndrome (ARDS) in geriatric patients following gastrointestinal perforation surgery, and constructed a model to validate its predictive value. METHODS: A retrospective analysis was conducted. The clinical data of geriatric patients (aged ≥ 60 years) after gastrointestinal perforation surgery admitted to the intensive care unit (ICU) of Hebei General Hospital from October 2017 to October 2024 were enrolled. Two groups were divided according to whether ARDS occurred postoperatively, and the differences in each index between the groups were compared. Lasso regression and multifactorial Logistic regression analyses were used to identify independent risk factors for the development of ARDS, and a prediction model was constructed based on these, which was presented using a nomogram. The receiver operator characteristic curve (ROC curve), calibration curve, and decision curve analysis (DCA) were plotted to evaluate the discrimination, accuracy, and clinical practicability of the model. RESULTS: A total of 155 geriatric patients following gastrointestinal perforation surgery were ultimately included in the analysis, among whom 43 developed ARDS, with an incidence rate of 27.7%. There were significantly differences in age, body mass index (BMI), acute kidney injury comorbidity, heart rate, onset time, the duration of surgery, the site of perforation, seroperitoneum, amount of bleeding, shock comorbidity, central venous pressure (CVP), C-reactive protein, and albumin between ARDS and non-ARDS groups. Lasso regression identified nine significant predictors: age, BMI, acute kidney injury comorbidity, onset time, seroperitoneum, shock comorbidity, CVP, hemoglobin, and albumin. Multivariate Logistic regression analysis identified BMI [odds ratio (OR) = 1.310, P < 0.001], hemoglobin (OR = 1.019, P = 0.045), seroperitoneum (OR = 1.001, P = 0.017), and albumin (OR = 0.871, P < 0.001) as independent risk factors for the occurrence of ARDS. A prediction model was constructed based on the above four independent risk factors, and the ROC curve showed that the area under the curve (AUC) of the model for predicting the occurrence of ARDS was 0.885 [95% confidence interval (95%CI) was 0.824-0.946], and internal validation was performed using bootstrap resampling (Bootstrap 500 times), which showed that the AUC value of the model was 0.886 (95%CI was 0.883-0.889). Calibration curves revealed excellent concordance between observed outcomes and model predictions. DCA indicated a high net benefit value for the model, which has good clinical utility. CONCLUSIONS BMI, hemoglobin, seroperitoneum, and albumin were identified as independent risk factors for ARDS in geriatric patients following gastrointestinal perforation surgery. The prediction model constructed using these four indicators facilitates early identification of high-risk individuals by clinicians.
Humans ; Respiratory Distress Syndrome/etiology* ; Retrospective Studies ; Aged ; Risk Factors ; Logistic Models ; Postoperative Complications ; Intestinal Perforation/surgery* ; Male ; ROC Curve ; Female ; Middle Aged ; Intensive Care Units ; Nomograms

OBJECTIVE: To identify the risk factors related to the timing of patients receiving extracorporeal membrane oxygenation (ECMO) initiation and construct a risk prediction model for ECMO initiation timing. METHODS: Patients who received ECMO admitted to the Second Xiangya Hospital of Central South University from January 2020 to January 2024 were retrospectively collected. The case data mainly included physiological and biochemical indicators 1 hour before ECMO initiation. According to the outcome of the patients, they were divided into survival group and death group. Univariate and multivariate Logistic regression analysis were used to analyze the predictors of mortality risk in patients with ECMO, and a nomogram prediction model was constructed. The discrimination, calibration accuracy, and goodness of the model were evaluated by the receiver operator characteristic curve (ROC curve), calibration curve, and the Hosmer-Lemeshow test, respectively. Decision curve analysis (DCA) evaluated the clinical net benefit rate of the model. RESULTS: A total of 81 ECMO patients were included, including 59 males and 22 females; age range from 16 to 61 years old, with a median age of 56.0 (39.5, 61.5) years old; 20 patients received veno-arterial (V-A) ECMO, and 61 patients received veno-venous (V-V) ECMO; 23 patients ultimately survived and 58 patients died. Univariate analysis showed that age, blood urea nitrogen, serum creatinine, D-dimer, arterial blood carbon dioxide partial pressure, and prothrombin time of the death group were all higher than those of the survival group, while albumin was slightly lower than that of the survival group. There was a statistically significant difference in the direct cause of ECMO initiation between the two groups. Multivariate Logistic regression analysis showed that age [odds ratio (OR) = 1.069, 95% confidence interval (95%CI) was 1.015-1.125, P = 0.012], direct cause of ECMO initiation [with heart failure as the reference, return of spontaneous circulation (ROSC) after cardiopulmonary support (OR = 30.672, 95%CI was 1.265-743.638, P = 0.035), novel coronavirus infection (OR = 8.666, 95%CI was 0.818-91.761, P = 0.073), other severe pneumonia (OR = 4.997, 95%CI was 0.558-44.765, P = 0.150)], pre-ECMO serum creatinine (OR = 1.008, 95%CI was 1.000-1.016, P = 0.044), prothrombin time (OR = 1.078, 95%CI was 0.948-1.226, P = 0.252), and D-dimer (OR = 1.135, 95%CI was 1.047-1.231, P = 0.002) were entered into the final regression equation. A nomogram prediction model was developed based on these five factors. The area under the ROC curve (AUC) of the model was 0.889 (95%CI was 0.819-0.959), higher than the AUC of the sequential organ failure assessment (SOFA; AUC = 0.604, 95%CI was 0.467-0.742). The calibration curve showed good consistency between the model predictions and the observed results. The Hosmer-Lemeshow goodness-of-fit test showed that χ ² = 4.668, P = 0.792. DCA analysis showed that when the risk threshold was 0-0.8, the net benefit rate was greater than 0, which was significantly better than that of SOFA score. CONCLUSIONS The risk prediction model for the timing of ECMO initiation, constructed using five factors (age, direct cause of ECMO initiation, thrombin time, serum creatinine, and D-dimer), demonstrated good discrimination and calibration. It can serve as a pre-initiation assessment tool to identify and predict post-initiation mortality risk in ECMO patients.
Humans ; Extracorporeal Membrane Oxygenation ; Middle Aged ; Male ; Female ; Retrospective Studies ; Adult ; Risk Factors ; Adolescent ; Young Adult ; Logistic Models ; Nomograms ; ROC Curve ; Time Factors ; Risk Assessment

OBJECTIVE: To observe change in serum growth differentiation factor 11 (GDF11) and killer cell lectin-like receptor B1 (KLRB1), to construct a nomogram model for 28-day death in patients with septic shock, and to explore its predictive value. METHODS: A prospective observational study was conducted. The patients with septic shock admitted to the emergency intensive care unit (ICU) of Cangzhou Central Hospital from September 2023 to March 2025 were selected as the septic shock group, the patients with sepsis admitted to the emergency general ward during the same period were selected as the sepsis group, and healthy individuals undergoing physical examination during the same period were selected as the control group. On the day of hospital admission or physical examination for the research subjects, the levels of serum GDF11 and KLRB1 were detected by enzyme-linked immunosorbent assay (ELISA). The patients with septic shock were divided into survival and death groups based on their 28-day survival status. The patients' gender, age, past medical history, infection site, severity of illness, mechanical ventilation, blood purification, infection indicators, biochemical indicators, coagulation function indicators, and blood lactic acid (Lac) were collected. The clinical data of the patients with septic shock between the two groups with different prognoses were compared. Multivariate Logistic regression analysis was used to screen the risk factors for 28-day death in patients with septic shock, and bivariate Pearson correlation analysis was conducted. A nomogram model was constructed based on the risk factors for 28-day death in patients with septic shock. The discrimination and calibration of the nomogram model were evaluated using the receiver operator characteristic curve (ROC curve), Hosmer-Lemeshow goodness-of-fit test, and calibration curve. The clinical utility of the model was evaluated using clinical decision curve analysis (DCA). RESULTS: A total of 168 patients in the emergency ICU were enrolled in the septic shock group, 40 patients in the emergency general ward were enrolled in the sepsis group, and 40 healthy individuals were enrolled in the control group. Compared with the healthy control group, the serum GDF11 levels in the sepsis and septic shock groups were significantly increased (μg/L: 13.09±3.51, 19.28±5.36 vs. 4.17±0.92, both P < 0.05), and the serum KLRB1 levels were significantly decreased (ng/L: 57.36±11.28, 45.52±9.07 vs. 84.19±17.16, both P < 0.05), with more significant changes in the septic shock group (both P < 0.05). Among the 168 patients with septic shock, 96 survived and 72 died within 28 days. Compared with the survival group, the serum GDF11 level in the death group was significantly increased (μg/L: 24.24±4.81 vs. 15.56±4.62, P < 0.05), and the serum KLRB1 level was significantly decreased (ng/L: 28.53±8.69 vs. 58.26±9.45, P < 0.05). There were also statistically significant differences in sequential organ failure assessment (SOFA) score, acute physiology and chronic health evaluation II (APACHEII) score, procalcitonin (PCT), activated partial thromboplastin time (APTT), D-dimer, and Lac between the two groups. Multivariate Logistic regression analysis showed that SOFA score [odds ratio (OR) = 1.96, 95% confidence interval (95%CI) was 1.38-3.65), Lac (OR = 1.38, 95%CI was 1.09-2.01), GDF11 (OR = 1.54, 95%CI was 1.21-2.33) and KLRB1 (OR = 0.64, 95%CI was 0.41-0.78) were independent risk factors for 28-day death in patients with septic shock (all P < 0.05). Bivariate Pearson correlation analysis showed that SOFA score was significantly positively correlated with Lac and GDF11 (r value was 0.37 and 0.58, respectively, both P < 0.05), and significantly negatively correlated with KLRB1 (r = -0.72, P < 0.05). A nomogram model was constructed based on the risk factors for 28-day death in patients with septic shock. ROC curve analysis showed that the area under the ROC curve (AUC) of the nomogram model for predicting 28-day death in patients with septic shock was 0.963 (95%CI was 0.929-0.990), indicating that the model had good discrimination and predictive ability. The Hosmer-Lemeshow goodness-of-fit test (χ ² = 9.578, P = 0.295) and calibration curve indicated that the predicted values of the model were in good agreement with the actual values. DCA indicated that the model provided a high net benefit for clinical decision-making. CONCLUSIONS The serum GDF11 level was significantly increased and the KLRB1 level was significantly decreased in patients with septic shock. The nomogram model based on GDF11 and KLRB1 could more accurately evaluate the 28-day death of patients with septic shock.
Humans ; Shock, Septic/blood* ; Nomograms ; Prospective Studies ; Prognosis ; Male ; Female ; Middle Aged ; Aged ; Intensive Care Units

OBJECTIVE: To construct machine learning prediction model for sepsis-associated encephalopathy (SAE), and analyze the application value of the model on early identification of SAE risk in elderly septic patients. METHODS: Patients aged over 60 years with a primary diagnosis of sepsis admitted to intensive care unit (ICU) from 2008 to 2023 were selected from Medical Information Mart for Intensive Care-IV 2.2 (MIMIC-IV 2.2). Demographic variables, disease severity scores, comorbidities, interventions, laboratory indicators, and hospitalization details were collected. Key factors associated with SAE were identified using univariate Logistic regression analysis. The data were randomly divided into training and validation sets in a 7 : 3 ratio. Multivariable Logistic regression analysis was conducted in the training set and visualized using a nomogram model for prediction of SAE. The discrimination of the model was evaluated in the validation set using the receiver operator characteristic curve (ROC curve), and its calibration was assessed using calibration curve. Furthermore, multiple machine learning algorithms, including multi-layer perceptron (MLP), support vector machine (SVM), naive bayes (NB), gradient boosting machine (GBM), random forest (RF), and extreme gradient boosting (XGB), were constructed in the training set. Their predictive performance was subsequently evaluated on the validation set. Taking the XGB model as an example, the interpretability of the model through the SHapley Additive exPlanations (SHAP) algorithm was enhanced to identify the key predictive factors and their contributions. RESULTS: A total of 2 204 septic patients were finally enrolled, of whom 840 developed SAE (38.1%). A total of 21 variables associated with SAE were screened through univariate Logistic regression analysis. Multivariable Logistic regression analysis showed that endotracheal intubation [odds ratio (OR) = 0.40, 95% confidence interval (95%CI) was 0.19-0.88, P < 0.001], oxygen therapy (OR = 0.76, 95%CI was 0.53-0.95, P = 0.023), tracheotomy (OR = 0.20, 95%CI was 0.07-0.53, P < 0.001), continuous renal replacement therapy (CRRT; OR = 0.32, 95%CI was 0.15-0.70, P < 0.001), cerebrovascular disease (OR = 0.31, 95%CI was 0.16-0.60, P < 0.001), rheumatic disease (OR = 0.44, 95%CI was 0.19-0.99, P < 0.001), male (OR = 0.68, 95%CI was 0.54-0.86, P = 0.001), and maximum anion gap (AG; OR = 0.95, 95%CI was 0.93-0.97, P < 0.001) were associated with an decreased probability of SAE, and age (OR = 1.05, 95%CI was 1.03-1.06, P < 0.001), acute physiology score III (APSIII; OR = 1.02, 95%CI was 1.01-1.02, P < 0.001), Oxford acute severity of illness score (OASIS; OR = 1.04, 95%CI was 1.03-1.06, P < 0.001), and length of hospital stay (OR = 1.01, 95%CI was 1.01-1.02, P < 0.001) were associated with an increased probability of SAE. A nomogram model was constructed based on these variables. In the validation set, ROC curve analysis showed that the model achieved an area under the ROC curve (AUC) of 0.723, and the calibration curve showed good consistency between the predicted probability of the model and the observed probability. Among the machine learning algorithms, including MLP, SVM, NB, GBM, RF, and XGB, the SVM model and RF model demonstrated relatively good predictive performance, with AUC of 0.748 and 0.739, respectively, and the sensitivity was both exceeding 85%. The predictive performance of the XGB model was explained through SHAP analysis, and the results indicated that APSIII score (SHAP value was 0.871), age (SHAP value was 0.521), and OASIS score (SHAP value was 0.443) were important factors affecting the predictive performance of the model. CONCLUSIONS The machine learning-based SAE prediction model exhibits good predictive capability and holds significant application value for the early identification of SAE risk in elderly septic patients.
Humans ; Machine Learning ; Aged ; Sepsis-Associated Encephalopathy ; Sepsis/complications* ; Intensive Care Units ; Logistic Models ; Middle Aged ; Male ; ROC Curve ; Female ; Bayes Theorem ; Nomograms ; Support Vector Machine ; Algorithms

OBJECTIVES: This study aimed to construct a risk prediction model for the occurrence of the demora-lization syndrome in patients with oral cancer and provide a scientific basis for the prevention of this syndrome in patients with oral cancer and the development of personalized care programs. METHODS: A total of 486 patients with oral cancer in West China Hospital of Stomatology of Sichuan University and Sun Yat-sen Memorial Hospital of Sun Yat-sen University from 2024 March to July were selected by convenience sampling. We integrated clinical data and evidence from previous studies to identify the key variables affecting the demoralization syndrome in patients with oral cancer. The 486 patients were divided into a training set and a validation set in an 8∶2 ratio. A clinical risk prediction model was established based on the individual data of 365 patients in the development cohort. Through least absolute shrinkage and selection operator (LASSO) regression, a moderate to severe risk prediction model of demoralization syndrome in oral cancer was constructed, and a clinical machine-learning nomogram was constructed. Bootstrap resampling was used for internal validation. The data of 121 patients in the validation cohort were externally validated. RESULTS: The incidence of the demoralization syndrome in patients with oral cancer was 405 cases (83.3%), of which 279 cases (57.4%) were mild, 176 cases (36.2%) were moderate, and 31 cases (6.4%) were severe. The core model, including patient education level, disease understanding, and MDASI-HN score, was used to predict the risk of outcome. Internal validation of the model yielded C statistic of 0.783 6 (95% CI: 0.78-0.87), beta of 0.843 4, and calibration intercept of -0.040 6. Through external validation, the validation set C statistic was 0.80 (95%CI: 0.71-0.87), beta was 0.80, and calibration intercept was -0.08. CONCLUSIONS Our risk prediction mo-del of the demoralization syndrome in patients with oral cancer performed robustly in validation cohorts of different nur-sing environments. The model has good correction and good discrimination and can be used as an evaluation and prediction item at admission.
Humans ; Mouth Neoplasms/complications* ; Male ; Female ; Nomograms ; Middle Aged ; Syndrome ; Aged ; Adult ; Risk Factors ; Risk Assessment ; Machine Learning