A nomogram for predicting the in-hospital mortality after large hemispheric infarction

Validation and refinement of a predictive nomogram using artificial intelligence: assessing in-hospital mortality in patients with large hemispheric cerebral infarction

Background

Large Hemispheric Infarction (LHI) poses significant mortality and morbidity risks, necessitating predictive models for in-hospital mortality. Previous studies have explored LHI progression to malignant cerebral edema (MCE) but have not comprehensively addressed in-hospital mortality risk, especially in non-decompressive hemicraniectomy (DHC) patients.

Methods

Demographic, clinical, risk factor, and laboratory data were gathered. The population was randomly divided into Development and Validation Groups at a 3:1 ratio, with no statistically significant differences observed. Variable selection utilized the Bonferroni-corrected Boruta technique (p < 0.01). Logistic Regression retained essential variables, leading to the development of a nomogram. ROC and DCA curves were generated, and calibration was conducted based on the Validation Group.

Results

This study included 314 patients with acute anterior-circulating LHI, with 29.6% in the Death group (n = 93). Significant variables, including Glasgow Coma Score, Collateral Score, NLR, Ventilation, Non-MCA territorial involvement, and Midline Shift, were identified through the Boruta algorithm. The final Logistic Regression model led to a nomogram creation, exhibiting excellent discriminative capacity. Calibration curves in the Validation Group showed a high degree of conformity with actual observations. DCA curve analysis indicated substantial clinical net benefit within the 5 to 85% threshold range.

Conclusion

We have utilized NIHSS score, Collateral Score, NLR, mechanical ventilation, non-MCA territorial involvement, and midline shift to develop a highly accurate, user-friendly nomogram for predicting in-hospital mortality in LHI patients. This nomogram serves as valuable reference material for future studies on LHI patient prognosis and mortality prevention, while addressing previous research limitations.

Epidemiology and Nomogram for Predicting the Cancer-specific Survival of Ovarian Granulosa Cell Tumor: A Seer database study

OBJECTIVE

ovarian granulosa cell tumor (OGCT) is a kind of infrequent ovarian malignant tumor with limited epidemiological data available. we established a predictive nomograph to verify the clinical prognosis.

METHODS

1005 diagnosed with ovarian granulosa cell tumor (OGCT) were extracted from Surveillance, Epidemiology, and End Results (SEER) public database from 2000-2018. Kaplan-Meier analysis was applied to distinguish risk factors, univariate and multivariate Cox analyses were used to determine the independent prognostic factors for cancer-specific survival (CSS) of OGCT patients. The obtained prognostic variables were combined to construct a nomogram model for predicting CSS in OGCT patients.

RESULTS

Model performance was detected and evaluated with ROC curves and calibration plots. Data collected from 1005 patients were divided into two groups: training cohort(n=703,70%) and validation cohort(n=302,30%). The multivariate Cox model identified five covariates including age, marital status, AJCC stages, surgery and chemotherapy as independent interfering factors of CSS. The nomogram has shown a promising and excellent accuracy in evaluating 3 -, 5 -, 8-year CSS in OGCT patients. In terms of the CSS of the training cohort, the AUC values of the 3 -, 5 -, 8-year ROC curves were 0.819,0.8,0.819, while in terms of the CSS of the validation cohort, the AUC values of the validation cohort were 0.822,0.84,0.823, respectively. All the calibration curves showed pleasant consistency between predicted and actual survival rates. The nomogram model established in the study can improve the veracity of prognosis prediction, thereby improving the accuracy of individualized survival risk assessment, and providing targeted and constructive recommendations for specific treatment options.

CONCLUSION

Age, advanced clinical stage, widower and without surgery therapy are independent risk factors for poor prognosis and the nomogram we constructed can help clinicians efficiently recognize high-risk OGCT patients to guide targeted therapies and improve their outcomes.

Interpretable machine learning models for predicting in-hospital death in patients in the intensive care unit with cerebral infarction

BACKGROUND

Research on patients with cerebral infarction in the Intensive Care Unit (ICU) is still lacking. Our study aims to develop and validate multiple machine-learning (ML) models using two large ICU databases-Medical Information Mart for Intensive Care version III (MIMIC-III) and eICU Research Institute Database (eRI)-to guide clinical practice.

METHODS

We collected clinical data from patients with cerebral infarction in the MIMIC-III and eRI databases within 24 h of admission. The opinion of neurologists and the Least Absolute Shrinkage and Selection Operator regression was used to screen for relevant clinical features. Using eRI as the training set and MIMIC-III as the test set, we developed and validated six ML models. Based on the results of the model validation, we select the best model and perform the interpretability analysis on it.

RESULTS

A total of 4,338 patients were included in the study (eRI:3002, MIMIC-III:1336), resulting in a total of 18 clinical characteristics through screening. Model validation results showed that random forest (RF) was the best model, with AUC and F1 scores of 0.799 and 0.417 in internal validation and 0.733 and 0.498 in external validation, respectively; moreover, its sensitivity and recall were the highest of the six algorithms for both the internal and external validation. The explanatory analysis of the model showed that the three most important variables in the RF model were Acute Physiology Score-III, Glasgow Coma Scale score, and heart rate, and that the influence of each variable on the judgement of the model was consistent with medical knowledge.

CONCLUSION

Based on a large sample of patients and advanced algorithms, our study bridges the limitations of studies on this area. With our model, physicians can use the admission information of cerebral infarction patients in the ICU to identify high-risk groups among them who are prone to in-hospital death, so that they could be more alert to this group of patients and upgrade medical measures early to minimize the mortality of patients.

Prognostic Value of Neutrophil-to-Lymphocyte Ratio in Stroke: A Systematic Review and Meta-Analysis

Background: Stroke has become a major problem around the world, which is one of the main causes of long-term disability. Therefore, it is important to seek a biomarker to predict the prognosis of patients with stroke. This meta-analysis aims to clarify the relationship between the neutrophil-to-lymphocyte ratio (NLR) and the prognosis of stroke patients.

Methods: This study was pre-registered in PROSPERO (CRD42020186544). We performed systematic research in PubMed, Web of Science, and EMBASE databases for studies investigating the prognostic value of NLR. Based on the enrolled studies, patients were divided into the low-NLR cohort and the high-NLR cohort. Odds ratios (ORs) with 95% confidence intervals (CIs) were extracted and analyzed by the Review Manager 5.3 and Stata 12.0 software. Heterogeneity was estimated by using Cochran's Q test and I² value. Sensitivity analyses and subgroup analyses were also performed to explore the potential sources of heterogeneity. Publication bias was assessed with funnel plots and assessed by Egger's tests.

Results: Forty-one studies with 27,124 patients were included. In the overall analysis, elevated NLR was associated with an increased mortality in acute ischemic stroke (AIS) patients (OR = 1.12, 95% CI = 1.07–1.16) and in acute hemorrhagic stroke (AHS) patients (OR = 1.23, 95% CI = 1.09–1.39), poorer outcomes in AIS patients (OR = 1.29, 95% CI = 1.16–1.44), and in AHS patients (OR = 1.11, 95% CI = 1.03–1.20). While in terms of hemorrhagic transformation (HT), elevated NLR was associated with an increased incidence of HT in AIS patients (OR = 1.15, 95% CI = 1.08–1.23).

Conclusions: This study demonstrated that elevated NLR was significantly associated with poor prognosis of stroke patients. High NLR is associated with a 1.1- to 1.3-fold increased risk of poor outcomes of AIS/AHS patients. NLR could be helpful as a potential prognostic biomarker to guide clinical decision making.

Systematic Review Registration:https://www.crd.york.ac.uk/prospero/display_record.php?ID=CRD42020186544.

Developing a Nomogram to Predict the Probability of Subsequent Vascular Events at 6-Month in Chinese Patients with Minor Ischemic Stroke

Purpose

To develop a nomogram to predict the risk of subsequent vascular events (SVE) at 6-month in Chinese patients with minor ischemic stroke (MIS).

Patients and Methods

We performed a retrospective analysis of 260 MIS patients, which were randomly divided into a derivation set (193 cases) and a verification set (67 cases) at a ratio of 3:1. Multi-factor logistic regression was used to construct a predictive model of SVE from the derivation set and verify it in the verification set.

Results

Finally, there were 51 cases (19.6%) of SVE in 260 MIS cases. Age, fasting blood glucose, metabolic syndrome, number of lesions found on MRI, and the infarct size were used to construct the prediction model and nomogram. The AUC in the derivation set was 0.901, with a sensitivity of 0.795, a specificity of 0.877, a positive likelihood ratio of 6.443, and a negative likelihood ratio of 0.234. The AUC in the verification set was 0.897, which was not significantly different from the derivation set (P = 0.937). The predictive model based on clinical parameters has good diagnostic efficiency and robustness.

Conclusion

The nomogram can provide personalized predictions for the 6-month SVE risk in Chinese MIS patients.

A nomogram to predict in-hospital mortality of neonates admitted to the intensive care unit

Background

To explore the influencing factors for in-hospital mortality in the neonatal intensive care unit (NICU) and to establish a predictive nomogram.

Methods

Neonatal data were extracted from the Medical Information Mart for Intensive Care III (MIMIC-III) database. Both univariate and multivariate logit binomial general linear models were used to analyse the factors influencing neonatal death. The area under the receiver operating characteristics (ROC) curve was used to assess the predictive model, which was visualized by a nomogram.

Results

A total of 1258 neonates from the NICU in the MIMIC-III database were eligible for the study, including 1194 surviving patients and 64 deaths. Multivariate analysis showed that red cell distribution width (RDW) (odds ratio [OR] 0.813, p=0.003) and total bilirubin (TBIL; OR 0.644, p<0.001) had protective effects on neonatal in-hospital death, while lymphocytes (OR 1.205, p=0.025), arterial partial pressure of carbon dioxide (PaCO₂; OR 1.294, p=0.016) and sequential organ failure assessment (SOFA) score (OR 1.483, p<0.001) were its independent risk factors. Based on this, the area under the curve of this predictive model was up to 0.865 (95% confidence interval 0.813 to 0.917), which was also confirmed by a nomogram.

Conclusions

The nomogram constructed suggests that RDW, TBIL, lymphocytes, PaCO₂ and SOFA score are all significant predictors for in-hospital mortality in the NICU.

A Prediction Nomogram for Acute Kidney Injury in Very-Low-Birth-Weight Infants: A Retrospective Study

Background and objective: Acute kidney injury (AKI) is recognized as an independent predictor for mortality in very-low-birth-weight (VLBW) infants and is reported to have a high incidence. In this study, we sought to identify the predictors for AKI in VLBW infants and thereby develop a prediction nomogram for the early detection and management of VLBW infants at high risk of developing AKI. Methods: We designed a retrospective study wherein we investigated the baseline hospitalization data of VLBW infants treated at our hospital between January 2012 and October 2018. Independent predictors of AKI in VLBW infants, as identified by multivariate logistic regression, were incorporated into a model. Hosmer-Lemeshow test was used to test the goodness of fit of the model, and a receiver operating characteristic (ROC) curve was plotted to assess the discriminative ability of the model. The model was internally validated using the 10-fold cross-validation method. A nomogram was plotted to predict the risk of AKI in VLBW infants on the basis of the results of multivariate logistic regression analysis. Results: We investigated the data of 604 VLBW infants, of which 144 (23.8%) developed AKI; in 111 (77.1%) of these infants, AKI occurred within 7 days of birth. Multivariate logistic regression analysis identified the following as predictive factors for AKI in VLBW infants: gestational age, red blood cell count within 3 days of birth, serum calcium concentration within 3 days of birth, maternal age of ≥35 years, and pulmonary arterial hypertension or myocardial injury. Furthermore, the nomogram was found to be effective in estimating the risk of AKI in VLBW infants, with an area under the curve (AUC) of 0.794 [95% confidence interval (CI): 0.754-0.834; P < 0.001]. Internal validation done by cross-validation showed that the average AUC was 0.788. Conclusion: The nomogram developed in this study was found to be sensitive and specific for the preoperative prediction of AKI in VLBW infants, as per the Kidney Disease: Improving Global Outcomes (KDIGO) criteria modified for neonates.