Understanding the complexity of sepsis mortality prediction via rule discovery and analysis: a pilot study

Machine learning-based prognostic model for 30-day mortality prediction in Sepsis-3

BACKGROUND

Sepsis poses a critical threat to hospitalized patients, particularly those in the Intensive Care Unit (ICU). Rapid identification of Sepsis is crucial for improving survival rates. Machine learning techniques offer advantages over traditional methods for predicting outcomes. This study aimed to develop a prognostic model using a Stacking-based Meta-Classifier to predict 30-day mortality risks in Sepsis-3 patients from the MIMIC-III database.

METHODS

A cohort of 4,240 Sepsis-3 patients was analyzed, with 783 experiencing 30-day mortality and 3,457 surviving. Fifteen biomarkers were selected using feature ranking methods, including Extreme Gradient Boosting (XGBoost), Random Forest, and Extra Tree, and the Logistic Regression (LR) model was used to assess their individual predictability with a fivefold cross-validation approach for the validation of the prediction. The dataset was balanced using the SMOTE-TOMEK LINK technique, and a stacking-based meta-classifier was used for 30-day mortality prediction. The SHapley Additive explanations analysis was performed to explain the model's prediction.

RESULTS

Using the LR classifier, the model achieved an area under the curve or AUC score of 0.99. A nomogram provided clinical insights into the biomarkers' significance. The stacked meta-learner, LR classifier exhibited the best performance with 95.52% accuracy, 95.79% precision, 95.52% recall, 93.65% specificity, and a 95.60% F1-score.

CONCLUSIONS

In conjunction with the nomogram, the proposed stacking classifier model effectively predicted 30-day mortality in Sepsis patients. This approach holds promise for early intervention and improved outcomes in treating Sepsis cases.

Can We Improve Mortality Prediction in Patients with Sepsis in the Emergency Department?

Background and Objectives: Sepsis represents a global health challenge and requires advanced diagnostic and prognostic approaches due to its elevated rate of morbidity and fatality. Our study aimed to assess the value of a novel set of six biomarkers combined with severity scores in predicting 28 day mortality among patients presenting with sepsis in the Emergency Department (ED). Materials and Methods: This single-center, observational, prospective cohort included sixty-seven consecutive patients with septic shock and sepsis enrolled from November 2020 to December 2022, categorized into survival and non-survival groups based on outcomes. The following were assessed: procalcitonin (PCT), soluble Triggering Receptor Expressed on Myeloid Cells-1 (sTREM-1), the soluble form of the urokinase plasminogen activator receptor (suPAR), high-sensitivity C-reactive protein (hs-CRP), interleukin-6 (IL-6), and azurocidin 1 (AZU1), alongside clinical scores such as the Quick Sequential Organ Failure Assessment (qSOFA), Systemic Inflammatory Response Syndrome (SIRS), the Sequential Organ Failure Assessment (SOFA), the Acute Physiology and Chronic Health Evaluation II (APACHE II), the Simplified Acute Physiology Score II and III (SAPS II/III), the National Early Warning Score (NEWS), Mortality in Emergency Department Sepsis (MEDS), the Charlson Comorbidity Index (CCI), and the Glasgow Coma Scale (GCS). The ability of each biomarker and clinical score and their combinations to predict 28 day mortality were evaluated. Results: The overall mortality was 49.25%. Mechanical ventilation was associated with a higher mortality rate. The levels of IL-6 were significantly higher in the non-survival group and had higher AUC values compared to the other biomarkers. The GCS, SOFA, APACHEII, and SAPS II/III showed superior predictive ability. Combining IL-6 with suPAR, AZU1, and clinical scores SOFA, APACHE II, and SAPS II enhanced prediction accuracy compared with individual biomarkers. Conclusion: In our study, IL-6 and SAPS II/III were the most accurate predictors of 28 day mortality for sepsis patients in the ED.

ChatGPT Predicts In-Hospital All-Cause Mortality for Sepsis: In-Context Learning with the Korean Sepsis Alliance Database

OBJECTIVES

Sepsis is a leading global cause of mortality, and predicting its outcomes is vital for improving patient care. This study explored the capabilities of ChatGPT, a state-of-the-art natural language processing model, in predicting in-hospital mortality for sepsis patients.

METHODS

This study utilized data from the Korean Sepsis Alliance (KSA) database, collected between 2019 and 2021, focusing on adult intensive care unit (ICU) patients and aiming to determine whether ChatGPT could predict all-cause mortality after ICU admission at 7 and 30 days. Structured prompts enabled ChatGPT to engage in in-context learning, with the number of patient examples varying from zero to six. The predictive capabilities of ChatGPT-3.5-turbo and ChatGPT-4 were then compared against a gradient boosting model (GBM) using various performance metrics.

RESULTS

From the KSA database, 4,786 patients formed the 7-day mortality prediction dataset, of whom 718 died, and 4,025 patients formed the 30-day dataset, with 1,368 deaths. Age and clinical markers (e.g., Sequential Organ Failure Assessment score and lactic acid levels) showed significant differences between survivors and non-survivors in both datasets. For 7-day mortality predictions, the area under the receiver operating characteristic curve (AUROC) was 0.70-0.83 for GPT-4, 0.51-0.70 for GPT-3.5, and 0.79 for GBM. The AUROC for 30-day mortality was 0.51-0.59 for GPT-4, 0.47-0.57 for GPT-3.5, and 0.76 for GBM. Zero-shot predictions using GPT-4 for mortality from ICU admission to day 30 showed AUROCs from the mid-0.60s to 0.75 for GPT-4 and mainly from 0.47 to 0.63 for GPT-3.5.

CONCLUSIONS

GPT-4 demonstrated potential in predicting short-term in-hospital mortality, although its performance varied across different evaluation metrics.

The relationship between potassium levels and 28-day mortality in sepsis patients: Secondary data analysis using the MIMIC-IV database

Objective

The goal of the research is to investigate the link between serum potassium levels and death after 28 days in sepsis patients, utilizing an extensive sample of patients from the multi-center Medical Information Mart for Intensive Care IV (MIMIC-IV) database. Current research on serum potassium levels and 28-day mortality in sepsis patients is questionable. This study adds to the growing body of evidence linking serum potassium levels to the 28-day possibility of death in patients with sepsis.

Methods

We collected 349,08 patients with sepsis from the retrospective cohort MIMIC-IV database, using serum potassium level on the first day of admission to the intensive care unit as the exposure variable and mortality at 28 days as the outcome variable. And controlled for confounding characteristics including gender, age, ethnicity, and vital signs during admission.

Results

Serum potassium has a U-shaped connection with 28-day mortality in patients suffering from sepsis. The turning point was 4.10 mmol/L (95 % confidence interval: 4.03 to 4.22). Serum potassium and 28-day mortality were negatively linked on the inflection point's left side (OR: 0.72; 95 % CI: 0.63 to 0.83, P < 0.0001); on the opposing side of the point of inflexion, serum potassium was enthusiastically attached to 28-day mortality. (OR: 1.13; 95 % CI: 1.06 to 1.21, P < 0.0001).

Conclusion

The research conducted found that too high or too low potassium levels were linked to a 28-day risk of mortality in humans with sepsis.

Prediction model of death risk in patients with sepsis and screening of biomarkers for prognosis of patients with myocardial injury

This study aimed to create a robust prediction model for sepsis patient mortality and identify key biomarkers in those with myocardial injury. A retrospective analysis of 261 sepsis inpatients was conducted, with 44 deaths and 217 recoveries. Key factors were assessed via univariate and multivariate analyses, revealing myocardial injury, shock, and pulmonary infection as independent mortality risk factors. Using LASSO regression, a reliable prediction model was developed and internally validated. Additionally, procalcitonin (PCT) emerged as a sensitive biomarker for myocardial injury prediction in sepsis patients. In summary, this study highlights myocardial injury, shock, and pulmonary infection as independent risk factors for sepsis-related deaths. The LASSO-based prediction model effectively forecasts the prognosis of septic patients with myocardial injury, with PCT showing promise as a predictive biomarker.

A Rulefit-based prognostic analysis using structured MRI report to select potential beneficiaries from induction chemotherapy in advanced nasopharyngeal carcinoma: A dual-centre study

BACKGROUND AND PURPOSE

Structured MRI report facilitate prognostic prediction for nasopharyngeal carcinoma (NPC). However, the intrinsic association among structured variables is not fully utilised. This study aimed to investigate the performance of a Rulefit-based model in feature integration behind structured MRI report and prognostic prediction in advanced NPC.

MATERIALS AND METHODS

We retrospectively enrolled 1207 patients diagnosed with non-metastatic advanced NPC from two centres, and divided into training (N = 544), internal testing (N = 367), and external testing (N = 296) cohorts. Machine learning algorithms including multivariate analysis, deep learning, Lasso, and Rulefit were used to establish corresponding prognostic models. The concordance indices (C- indices) of three clinical and six combined models with different algorithms for overall survival (OS) prediction were compared. Survival benefits of induction chemotherapy (IC) were calculated among risk groups stratified by different models. A website was established for individualised survival visualisation.

RESULTS

Incorporating structured variables into Stage model significantly improved the prognostic prediction performance. Six prognostic rules with structured variables were identified by Rulefit. OS prediction of Rules model was comparable to Lasso model in internal testing cohort (C-index: 0.720 vs. 0.713, P = 0.100) and achieved the highest C-index of 0.711 in external testing cohort, indicating better generalisability. The Rules model stratified patients into risk groups with significant 5-year OS differences in each cohort, and revealed significant survival benefits from additional IC in high-risk group.

CONCLUSION

The Rulefit-based Rules model, with the revelation of intrinsic associations behind structured variables, is promising in risk stratification and guiding individualised IC treatment for advanced NPC.

Construction and efficacy evaluation of an early warning scoring system for septic shock in patients with digestive tract perforation: A retrospective cohort study

Objective

To establish an early warning scoring system for septic shock in patients with digestive tract perforation (DTP) and evaluate its diagnostic efficacy.

Methods

Patients with surgically confirmed or clinically diagnosed DTP admitted to the Department of Intensive Care Medicine of Fujian Provincial Hospital from June 2012 to October 2021 were retrospectively analyzed. General demographic characteristics, perforation-related information, vital signs, common laboratory indicators, and common ICU scores (Glasgow Coma Scale score, Acute Physiology and Chronic Health Evaluation-II score,Sequential Organ Failure Assessment score) were collected. The patients were divided into shock group and non-shock group according to whether the patients had septic shock during hospitalization. The risk factors of septic shock were screened by basic statistical analysis and multivariate Logistic regression analysis. The receiver operating characteristic curve was drawn to determine the cut-off value of the continuous indicators and discretized with reference to clinic, and the corresponding score was set according to the β regression coefficient of each variable.

Results

A total of 176 patients with DTP were included. The average age of the patients was 64.13 ± 14.67 years old, and 74.40% were males. The incidence of septic shock was 30.11% (53/176). Multivariate Logistic regression analysis showed that the highest heart rate≥105 beats/min, Glasgow Coma Scale score≤14 points, lactic acid≥5.75 mmol/L, procalcitonin≥41.47 ug/L, C-reactive protein≥222.5 mg/L were independent risk factors for septic shock in patients with DTP. The total score of clinical diagnostic scoring system of septic shock in patients with DTP was 6 points, including the highest heart rate≥105 beats/min (1 point), lactic acid≥5.75 mmol/L (two points), procalcitonin≥41.47 ug/L (one point), C-reactive protein≥222.5 mg/L (1 point), and Glasgow Coma Scale score≤14 points (1 point). The area under ROC curve (AUC) of this scoring system was 0.789 and the 95% confidence interval was 0.717–0.860 (P &lt; 0.001); when the optimal cut-off value was 2.5, the sensitivity and specificity were 54.70 and 87.80%, respectively.

Conclusion

This new score system has its certain clinical value and has important guiding significance for clinicians to judge the prognosis of patients with DTP in time.