Related Experiment Video
Updated: Jun 20, 2025
Cefoperazone-treated Mouse Model of Clinically-relevant Clostridium difficile Strain R20291
Published on: December 10, 2016
Predicting Clostridioides difficile infection outcomes with explainable machine learning
Gregory R Madden1, Rachel H Boone2, Emmanuel Lee3
1Division of Infectious Diseases & International Health, Department of Medicine, University of Virginia School of Medicine, Charlottesville, VA, USA; Office of Hospital Epidemiology/Infection Prevention & Control, University of Virginia School of Medicine, Charlottesville, VA, USA.
Background:
Clostridioides difficile infection results in life-threatening short-term outcomes and the potential for subsequent recurrent infection. Predicting these outcomes at diagnosis, when important clinical decisions need to be made, has proven to be a difficult task.
Methods:
52 clinical features from existing models or the literature were collected retrospectively within ±48 h of diagnosis among 1660 inpatient infections. A modified desirability of outcome ranking (DOOR) was designed to encompass clinically-important severe events attributable to the acute infection (intensive care transfer due to sepsis, shock, colectomy/ileostomy, mortality) and/or 60-day recurrence. A deep neural network was constructed and interpreted using SHapley Additive exPlanations (SHAP). High-importance features were used to train a reduced, shallow network and performance was compared to existing conventional models (7 severity, 7 recurrence; after summing DOOR probabilities to align with conventional binary outputs) using area under the ROC curve (AUROC) and DeLong tests.
Findings:
The full (52-feature) model achieved an out-of-sample AUROC 0.823 for severity and 0.678 for recurrence. SHAP identified 13 unique, highly-important features (age, hypotension, initial treatment, onset, PCR cycle threshold, number of prior episodes, antibiotic exposure, fever, hypotension, pressors, leukocytosis, creatinine, lactate) that were used to train a reduced model, which performed similarly to the full model (severity AUROC difference P = 0.130; recurrence P = 0.426) and significantly better than the top severity model (reduced model predicting severity 0.837, ATLAS 0.749; P = 0.001). The reduced model also outperformed the top recurrence model, but this was not statistically-significant (reduced model recurrence AUROC 0.653, IDSA Recurrence Risk Criteria 0.595; P = 0.196). The final, reduced model was deployed as a web application with real-time SHAP explanations.
Interpretation:
Our final model outperformed existing severity and recurrence models; however, it requires external validation. A DOOR output allows specific clinical questions to be asked with explainable predictions that can be feasibly implemented with limited computing resources.
Funding:
National Institutes of Health-Institute of Allergy and Infectious Diseases.
Insights
This study developed a new model to predict severe outcomes and recurrence of Clostridioides difficile infection. The model shows improved accuracy over existing methods, aiding clinical decision-making at diagnosis.
Area of Science:
- Medical Informatics
- Infectious Diseases
- Machine Learning
Background:
- Clostridioides difficile infection (CDI) presents significant short-term risks and potential for recurrence.
- Predicting CDI outcomes at diagnosis is challenging but crucial for clinical decision-making.
Purpose of the Study:
- To develop and validate a predictive model for severe outcomes and recurrence of CDI.
- To identify key clinical features for accurate CDI prognostication.
Main Methods:
- Retrospective collection of 52 clinical features from 1660 inpatient CDI cases.
- Development of a modified desirability of outcome ranking (DOOR) model using deep neural networks and SHAPley Additive exPlanations (SHAP).
- Comparison of model performance against existing severity and recurrence prediction models using AUROC.
Main Results:
- The full 52-feature model achieved AUROCs of 0.823 for severity and 0.678 for recurrence.
- SHAP identified 13 high-importance features, enabling a reduced model with similar performance.
- The reduced model significantly outperformed the top existing severity model (AUROC 0.837 vs. 0.749).
Conclusions:
- The developed model demonstrates superior performance in predicting CDI severity compared to existing tools.
- The model requires external validation but offers explainable predictions for clinical implementation.
- A web application with real-time SHAP explanations was developed for feasible use.
More Related Videos
04:09Predicting Treatment Response to Image-Guided Therapies Using Machine Learning: An Example for Trans-Arterial Treatment of Hepatocellular Carcinoma
Published on: October 10, 2018
07:15Machine Learning Algorithms for Early Detection of Bone Metastases in an Experimental Rat Model
Published on: August 16, 2020
Related Concept Videos
01:18Steps in Outbreak Investigation
01:15Receiver Operating Characteristic Plot