Related Experiment Video
Updated: Aug 12, 2026
A Data-Driven Approach to Quantifying Immune States in Sepsis
Published on: February 7, 2025
Utilizing Machine Learning for Diagnostic Assistance of Pediatric Sepsis and Septic Shock in Resource-Limited
Kaden Bunch1, Shamsun Nahar Shaima2, Gazi Md Salahuddin Mamun2
1Warren Alpert Medical School of Brown University, Providence, RI 02903, USA.
Insights
Machine learning models show promise for detecting pediatric sepsis and septic shock using simplified clinical data in resource-limited settings. The non-laboratory model performed comparably to those using lab data, but external validation is crucial.
Area of Science:
- Pediatric critical care
- Medical artificial intelligence
- Global health
Background:
- Sepsis is a major cause of child mortality globally, especially in low- and middle-income countries (LMICs).
- Diagnosing pediatric sepsis is challenging in LMICs due to limited healthcare resources and difficulties in timely recognition.
- This study addresses the need for practical diagnostic tools in resource-limited settings.
Purpose of the Study:
- To develop and evaluate machine learning (ML) models for detecting pediatric sepsis and septic shock.
- To utilize a simplified set of clinical data suitable for resource-limited environments.
- To assess the diagnostic performance of models with and without laboratory variables.
Main Methods:
- Secondary analysis of an observational study involving 100 children with suspected sepsis in Bangladesh.
- Developed ML models using clinical variables only, and combined clinical and laboratory variables.
- Assessed model performance using area under the precision-recall curve (AUPRC) and area under the receiver operating characteristic curve (AUROC) with 5-fold cross-validation.
Main Results:
- The non-laboratory ML model for sepsis achieved an AUPRC of 0.942 and AUROC of 0.945.
- Performance was comparable to the model including laboratory variables.
- Key predictors included SpO2:FiO2 ratio, Glasgow Coma Scale (GCS), and systolic blood pressure; septic shock detection showed lower AUROCs.
Conclusions:
- ML models using pragmatic clinical data show preliminary diagnostic capability for pediatric sepsis.
- A non-laboratory model demonstrated performance comparable to models using laboratory data.
- External validation in larger cohorts is essential before clinical implementation, especially for septic shock detection.
Background:
Sepsis is a leading cause of pediatric mortality worldwide, disproportionately affecting children in low- and middle-income countries (LMICs). However, timely recognition of potential sepsis and access to healthcare resources needed to diagnose pediatric sepsis according to international guidelines are challenging in LMICs. This exploratory study aimed to develop machine learning (ML) models to detect pediatric sepsis and septic shock using a simplified set of clinical data contextualized for practical use in resource-limited settings.
Methods:
This was a secondary analysis of an observational study of 100 children with potential sepsis admitted to a non-profit referral hospital in Dhaka, Bangladesh. The outcomes were sepsis as defined by a Phoenix Sepsis Score (PSS) ≥ 2 and septic shock (sepsis plus PSS cardiovascular sub-score ≥ 1). Models were trained using either clinical + laboratory variables or clinical-only variables. A single 24 h worst-value assessment window was derived per patient; stratified 5-fold cross-validation was used to maintain class proportions across the training and test folds. Model performance was assessed using area under the precision-recall curve (AUPRC) and area under the receiver operating characteristic curve (AUROC) with 95% confidence intervals (CIs) derived from a 2000-resample patient-level bootstrap of out-of-fold classifications. Logistic regression coefficients were used to assess feature contributions.
Results:
For sepsis classification, the non-laboratory model achieved an AUPRC of 0.942 (95% CI: 0.884-0.979) and an AUROC of 0.945 (95% CI: 0.890-0.983), with comparable performance from the clinical + laboratory model (AUPRC 0.941, 95% CI: 0.880-0.981; AUROC 0.945, 95% CI: 0.881-0.986). For septic shock, AUROCs of 0.870 (95% CI: 0.761-0.952) and 0.878 (95% CI: 0.758-0.967) were observed. However, these estimates should be interpreted cautiously, given the low prevalence (23%) and absence of external validation. SpO2:FiO2 ratio, GCS, and systolic blood pressure were consistently strong predictors across models.
Conclusions:
ML models using pragmatic clinical variables demonstrate preliminary diagnostic performance, with the non-laboratory model showing discrimination comparable to models incorporating laboratory data. Logistic regression demonstrated the most stable performance and may represent an early proof of concept for assistive diagnostic support. However, these models are not clinically usable without external validation. These findings are hypothesis-generating; external validation in larger, independent cohorts is essential before any clinical use, particularly for septic shock.