A predictive model for Epstein-Barr virus-associated hemophagocytic lymphohistiocytosis.
Frontiers in immunology
Background:Epstein-Barr virus-associated hemophagocytic lymphohistiocytosis (EBV-HLH) is a severe hyperinflammatory disorder induced by overactivation of macrophages and T cells. This study aims to identify the risk factors for the progression from infectious mononucleosis (EBV-IM) to EBV-HLH, by analyzing the laboratory parameters of patients with EBV-IM and EBV-HLH and constructing a clinical prediction model. The outcome of this study carries important clinical value for early diagnosis and treatment of EBV-HLH. Methods:A retrospective analysis was conducted on 60 patients diagnosed with EBV-HLH and 221 patients diagnosed with EBV-IM at our hospital between November 2018 and January 2024. Participants were randomly assigned to derivation and internal validation cohorts in a 7:3 ratio. LASSO regression and logistic regression analyses were employed to identify risk factors and construct the nomogram. Results:Ferritin (OR, 213.139; 95% CI, 8.604-5279.703; P=0.001), CD3CD16CD56% (OR, 0.011; 95% CI, 0-0.467; P=0.011), anti-EBV-NA-IgG (OR, 57.370; 95%CI, 2.976-1106.049; P=0.007), IL-6 (OR, 71.505; 95%CI, 2.118-2414.288; P=0.017), IL-10 (OR, 213.139; 95% CI, 8.604-5279.703; P=0.001) were identified as independent predictors of EBV-HLH. The prediction model demonstrated excellent discriminatory capability evidenced by an AUC of 0.997 (95% CI,0.993-1.000). When visualized using a nomogram, the ROC curves for the derivation and validation cohorts exhibited AUCs of 0.997 and 0.993, respectively. These results suggested that the model was highly stable and accurate. Furthermore, calibration curves and clinical decision curves indicated that the model possessed good calibration and offered significant clinical benefits. Conclusions:The nomogram, which was based on these five predictors, exhibited robust predictive value and stability, thereby can be used to aid clinicians in the early detection of EBV-HLH.
10.3389/fimmu.2024.1503118
Predictive modeling of ICU-AW inflammatory factors based on machine learning.
BMC neurology
BACKGROUND:ICU-acquired weakness (ICU-AW) is a common complication among ICU patients. We used machine learning techniques to construct an ICU-AW inflammatory factor prediction model to predict the risk of disease development and reduce the incidence of ICU-AW. METHODS:The least absolute shrinkage and selection operator (LASSO) technique was used to screen key variables related to ICU-AW. Eleven indicators, such as the presence of sepsis, glucocorticoids (GC), neuromuscular blocking agents (NBAs), length of ICU stay, Acute Physiology and Chronic Health Evaluation (APACHE II) II score, and the levels of albumin (ALB), lactate (LAC), glucose (GLU), interleukin-1β (IL-1β), interleukin-6 (IL-6), and interleukin-10 (IL-10), were used as variables to establish the prediction model. We divided the data into a dataset that included inflammatory factors and a dataset that excluded inflammatory factors. Specifically, 70% of the participants in both datasets were used as the training set, and 30% of the participants were used as the test set. Three machine learning methods, logistic regression (LR), random forest (RF), and extreme gradient boosting (XGB), were used in the 70% participant training set to construct six different models, which were validated and evaluated in the remaining 30% of the participants as the test set. The optimal model was visualized for prediction using nomograms. RESULTS:The logistic regression model including the inflammatory factors demonstrated excellent performance on the test set, with an area under the curve (AUC) of 82.1% and the best calibration curve fit, outperforming the other five models. The optimal model is represented visually in the nomograms. CONCLUSION:This study used easily accessible clinical characteristics and laboratory data that can aid in early clinical recognition of ICU-AW. The inflammatory factors IL-1β, IL-6, and IL-10 have high value for predicting ICU-AW. TRIAL REGISTRATION:The trial was registered at the Chinese Clinical Trial Registry with the registration number ChiCTR2300077968.
10.1186/s12883-024-03981-w
Interleukin-6 and thyroid-stimulating hormone index predict plaque stability in carotid artery stenosis: analyses by lasso-logistic regression.
Frontiers in cardiovascular medicine
Objective:To develop and validate a new prediction model based on the Lass-logistic regression with inflammatory serologic markers for the assessment of carotid plaque stability, providing clinicians with a reliable tool for risk stratification and decision-making in the management of carotid artery disease. Methods:In this study, we retrospectively collected the data of the patients who underwent carotid endarterectomy (CEA) from 2019 to 2023 in Nanjing Drum Tower Hospital. Demographic characteristics, vascular risk factors, and the results of preoperative serum biochemistry were measured and collected. The risk factors for vulnerable carotid plaque were analyzed. A Lasso-logistic regression prediction model was developed and compared with traditional logistic regression models. The Akaike information criterion (AIC) and Bayesian information criterion (BIC) were used to evaluate the performance of three models. Results:A total of 131 patients were collected in this study, including 66 (50.4%) in the vulnerable plaque group and 65 (49.6%) in the stable plaque group. The final Lasso-logistic regression model included 4 features:IL-6, TSH, TSHI, and TT4RI; AIC = 161.6376, BIC = 176.0136, both lower than the all-variable logistic regression model (AIC = 181.0881, BIC = 261.5936), and the BIC was smaller than the stepwise logistic regression model (AIC = 154.024, BIC = 179.9007). Finally, the prediction model was constructed based on the variables screened by the Lasso regression, and the model had favorable discrimination and calibration. Conclusions:The noninvasive prediction model based on IL-6 and TSHI is a quantitative tool for predicting vulnerable carotid plaques. It has high diagnostic efficacy and is worth popularizing and applying.
10.3389/fcvm.2024.1484273
[Establishment and validation of a sepsis 28-day mortality prediction model based on the lactate dehydrogenase-to-albumin ratio in patients with sepsis].
Zhonghua wei zhong bing ji jiu yi xue
OBJECTIVE:To develop and validate a predictive model of 28-day mortality in sepsis based on lactate dehydrogenase-to-albumin ratio (LAR). METHODS:Sepsis patients diagnosed in the department of intensive care medicine of the First Affiliated Hospital of Soochow University from August 1, 2017 to September 1, 2022 were retrospective selected. Clinical data, laboratory indicators, disease severity scores [acute physiology and chronic health evaluation II (APACHE II), sequential organ failure assessment (SOFA)] were collected. Patients were divided into death group and survival group according to whether they died at 28 days, and the difference between the two groups was compared. The dataset was randomly divided into training set and validation set according to 7 : 3. Lasso regression method was used to screen the risk factors affecting the 28-day death of sepsis patients, and incorporating multivariate Logistic regression analysis (stepwise regression) were included, a prediction model was constructed based on the independent risk factors obtained, and a nomogram was drawn. The nomogram prediction model was established. Receiver operator characteristic curve (ROC curve) was drawn to analyze and evaluate the predictive efficacy of the model. Hosmer-Lemeshow test, calibration curve and decision curve analysis (DCA) were used to evaluate the accuracy and clinical practicability of the model, respectively. RESULTS:A total of 394 patients with sepsis were included, with 248 survivors and 146 non-survivors at 28 days. Compared with the survival group, the age, proportion of chronic obstructive pneumonia, respiratory rate, lactic acid, red blood cell distribution width, prothrombin time, activated partial thromboplastin time, alanine aminotransferase, aspartate aminotransferase, blood urea nitrogen, creatinine, blood potassium, blood phosphorus, LAR, SOFA score, and APACHE II score in the death group were significantly increased, while oxygenation index, monocyte count, platelet count, fibrinogen, total cholesterol, triglycerides, high-density lipoprotein, low-density lipoprotein, and blood calcium were significantly reduced. In the training set, LAR, age, oxygenation index, blood urea nitrogen, lactic acid, total cholesterol, fibrinogen, blood potassium and blood phosphorus were screened by Lasso regression. Multivariate Logistic regression analysis finally included LAR [odds ratio (OR) = 1.029, 95% confidence interval (95%CI) was 1.014-1.047, P < 0.001], age (OR = 1.023, 95%CI was 1.005-1.043, P = 0.012), lactic acid (OR = 1.089, 95%CI was 1.003-1.186, P = 0.043), oxygenation index (OR = 0.996, 95%CI was 0.993-0.998, P = 0.002), total cholesterol (OR = 0.662, 95%CI was 0.496-0.865, P = 0.003) and blood potassium (OR = 1.852, 95%CI was 1.169-2.996, P = 0.010). A total of 6 predictor variables were used to establish a prediction model. ROC curve showed that the area under the curve (AUC) of the model in the training set and validation set were 0.773 (95%CI was 0.715-0.831) and 0.793 (95%CI was 0.703-0.884), which was better than APACHE II score (AUC were 0.699 and 0.745) and SOFA score (AUC were 0.644 and 0.650), and the cut-off values were 0.421 and 0.309, the sensitivity were 62.4% and 82.2%, and the specificity were 82.2% and 68.9%, respectively. The results of Hosmer-Lemeshow test and calibration curve showed that the predicted results of the model were in good agreement with the actual clinical observation results, and the DCA showed that the model had good clinical application value. CONCLUSIONS:The prediction model based on LAR has a good predictive value for 28-day mortality in patients with sepsis and can guide clinical decision-making.
10.3760/cma.j.cn121430-20231012-00865
Construction and validation of a nomogram model for predicting diabetic peripheral neuropathy.
Frontiers in endocrinology
Objective:Diabetic peripheral neuropathy (DPN) is a chronic complication of diabetes that can potentially escalate into ulceration, amputation and other severe consequences. The aim of this study was to construct and validate a predictive nomogram model for assessing the risk of DPN development among diabetic patients, thereby facilitating the early identification of high-risk DPN individuals and mitigating the incidence of severe outcomes. Methods:1185 patients were included in this study from June 2020 to June 2023. All patients underwent peripheral nerve function assessments, of which 801 were diagnosed with DPN. Patients were randomly divided into a training set (n =711) and a validation set (n = 474) with a ratio of 6:4. The least absolute shrinkage and selection operator (LASSO) logistic regression analysis was performed to identify independent risk factors and develop a simple nomogram. Subsequently, the discrimination and clinical value of the nomogram was extensively validated using receiver operating characteristic (ROC) curves, calibration curves and clinical decision curve analyses (DCA). Results:Following LASSO regression analysis, a nomogram model for predicting the risk of DPN was eventually established based on 7 factors: age (OR = 1.02, 95%CI: 1.01 - 1.03), hip circumference (HC, OR = 0.94, 95%CI: 0.92 - 0.97), fasting plasma glucose (FPG, OR = 1.06, 95%CI: 1.01 - 1.11), fasting C-peptide (FCP, OR = 0.66, 95%CI: 0.56 - 0.77), 2 hour postprandial C-peptide (PCP, OR = 0.78, 95%CI: 0.72 - 0.84), albumin (ALB, OR = 0.90, 95%CI: 0.87 - 0.94) and blood urea nitrogen (BUN, OR = 1.08, 95%CI: 1.01 - 1.17). The areas under the curves (AUC) of the nomogram were 0.703 (95% CI 0.664-0.743) and 0.704 (95% CI 0.652-0.756) in the training and validation sets, respectively. The Hosmer-Lemeshow test and calibration curves revealed high consistency between the predicted and actual results of the nomogram. DCA demonstrated that the nomogram was valuable in clinical practice. Conclusions:The DPN nomogram prediction model, containing 7 significant variables, has exhibited excellent performance. Its generalization to clinical practice could potentially help in the early detection and prompt intervention for high-risk DPN patients.
10.3389/fendo.2024.1419115
Comprehensive predictive model for cerebral microbleeds: integrating clinical and biochemical markers.
Frontiers in neuroscience
Background:Cerebral Microbleeds (CMBs) serve as critical indicators of cerebral small vessel disease and are strongly associated with severe neurological disorders, including cognitive impairments, stroke, and dementia. Despite the importance of diagnosing and preventing CMBs, there is a significant lack of effective predictive tools in clinical settings, hindering comprehensive assessment and timely intervention. Objective:This study aims to develop a robust predictive model for CMBs by integrating a broad range of clinical and laboratory parameters, enhancing early diagnosis and risk stratification. Methods:We analyzed extensive data from 587 neurology inpatients using advanced statistical techniques, including Least Absolute Shrinkage and Selection Operator (LASSO) and logistic regression. Key predictive factors such as Albumin/Globulin ratio, gender, hypertension, homocysteine levels, Neutrophil to HDL Ratio (NHR), and history of stroke were evaluated. Model validation was performed through Receiver Operating Characteristic (ROC) curves and Decision Curve Analysis (DCA). Results:The model demonstrated strong predictive performance with significant clinical applicability. Key predictors identified include the Albumin/Globulin ratio, homocysteine levels, and NHR, among others. Validation metrics such as the area under the ROC curve (AUC) and decision curve analysis confirmed the model's utility in predicting CMBs, highlighting its potential for clinical implementation. Conclusion:The comprehensive predictive model developed in this study offers a significant advancement in the personalized management of patients at risk for CMBs. By addressing the gap in effective predictive tools, this model facilitates early diagnosis and targeted intervention, potentially reducing the incidence of stroke and cognitive impairments associated with cerebral microbleeds. Our findings advocate for a more nuanced approach to cerebrovascular disease management, emphasizing the importance of multi-factorial risk profiling.
10.3389/fnins.2024.1429088