Related Experiment Video
Updated: Apr 28, 2026
Tools for the Real-Time Assessment of a Pseudomonas aeruginosa Infection Model
Published on: April 6, 2021
Machine learning-based risk prediction model for sepsis development in patients with multidrug-resistant Pseudomonas
Chang Li1, Ting Shi2, Guanyu Xiao3
1Department of Medical Laboratory, The Affiliated Huai'an No. 1 People's Hospital of Nanjing Medical University, Huai'an, Jiangsu, China.
Background:
Multidrug-resistant Pseudomonas aeruginosa (MDR-PA) infections present a critical healthcare challenge, often progressing to sepsis with high mortality. Current prediction tools lack specificity for drug-resistant organisms, hindering the early identification of high-risk patients. This study aimed to develop and validate an interpretable machine learning (ML) model to predict sepsis development in patients with MDR-PA infections.
Methods:
We conducted a multicenter retrospective study analyzing 2,001 patients with laboratory-confirmed MDR-PA infections from two major medical centers between January 2019 and May 2025. The derivation cohort included 1,182 patients, while 819 patients from an independent center served as the external validation cohort. Feature selection was performed using a hybrid approach combining LASSO regression and support vector machine-recursive feature elimination (SVM-RFE). Seven ML algorithms were evaluated, with model interpretability enhanced via SHapley Additive exPlanations (SHAP). A web-based calculator was subsequently developed to facilitate clinical implementation.
Results:
The sepsis incidence was approximately 7% across cohorts. Feature selection identified six key predictors: calcium level, chronic obstructive pulmonary disease (COPD), red blood cell distribution width-standard deviation (RDW-SD), intra-abdominal infection, invasive catheters, and prior antibiotic exposure. The Random Forest model demonstrated superior performance, achieving an AUC of 1.000 in the SMOTE-balanced training set, 0.837 in internal validation, and 0.816 in external validation. SHAP analysis highlighted COPD and calcium levels as the most significant contributors to sepsis risk.
Conclusions:
This study presents the first interpretable ML model specifically tailored for predicting sepsis onset in patients with MDR-PA infections. By addressing the limitations of general sepsis scores, our validated model and accompanying web-based tool provide clinicians with a precise, visualizable decision-support system to optimize early intervention strategies.
Insights
This study developed an interpretable machine learning model to predict sepsis in multidrug-resistant Pseudomonas aeruginosa infections. The model accurately identifies high-risk patients, improving early intervention for this critical healthcare challenge.
Area of Science:
- Infectious Diseases
- Machine Learning
- Critical Care Medicine
Background:
- Multidrug-resistant Pseudomonas aeruginosa (MDR-PA) infections pose a significant threat, often leading to sepsis with high mortality rates.
- Existing sepsis prediction tools lack specificity for drug-resistant pathogens, delaying critical care for high-risk patients.
- There is a need for specialized tools to predict sepsis development in MDR-PA infected individuals.
Purpose of the Study:
- To develop and validate an interpretable machine learning (ML) model for predicting sepsis onset in patients with MDR-PA infections.
- To enhance early identification and intervention for patients at high risk of sepsis due to MDR-PA.
- To provide a clinically applicable decision-support tool for healthcare providers.
Main Methods:
- A multicenter retrospective study analyzed 2,001 patients with laboratory-confirmed MDR-PA infections.
- Feature selection utilized a hybrid LASSO regression and SVM-RFE approach.
- Seven ML algorithms were evaluated, with SHAP used for interpretability, and a web-based calculator developed.
Main Results:
- Sepsis incidence was approximately 7% across cohorts.
- Key predictors identified: calcium level, COPD, RDW-SD, intra-abdominal infection, invasive catheters, and prior antibiotic exposure.
- The Random Forest model achieved AUCs of 0.837 (internal) and 0.816 (external) validation, with SHAP highlighting COPD and calcium levels as significant risk factors.
Conclusions:
- This study introduces the first interpretable ML model specifically for predicting sepsis in MDR-PA infections.
- The validated model and web tool address limitations of general sepsis scores, offering precise decision support.
- The tool aims to optimize early intervention strategies for improved patient outcomes in MDR-PA sepsis.
Related Concept Videos
01:25Clinical Significance of Antibiotic Resistance
01:25Mechanism of Antibiotic Resistance in MRSA