Advancing Bloodstream Infection Prediction Using Explanable Artificial Intelligence Framework

Rajeev Bopche,Lise Tuset Gustad,Jan Egil Afset,Birgitta Ehrnström,Jan Kristian Damås,Øystein Nytrø

DOI: https://doi.org/10.1101/2024.04.10.24305614

2024-04-12

Abstract:Bloodstream infections (BSIs) represent a critical public health concern, primarily due to their rapid progression and severe implications such as sepsis and septic shock. This study introduces an innovative Explanable Artificial Intelligence (XAI) framework, leveraging historical electronic health records (EHRs) to enhance BSI prediction. Unlike traditional models that rely heavily on real-time clinical data, our XAI-based approach utilizes a comprehensive dataset incorporating demographic data, laboratory results, and full medical histories from St. Olavs Hospital, Trondheim, Norway, covering 35,591 patients between 2015 and 2020. We developed models to differentiate between high-risk and low-risk BSI cases effectively, optimizing healthcare resource allocation and potentially reducing healthcare costs. Our results demonstrate superior predictive accuracy, particularly the tree-based models, which significantly outperformed contemporary models in both specificity and sensitivity metrics.

Health Informatics

What problem does this paper attempt to address?

The problem this paper attempts to address is: how to use an explainable artificial intelligence (XAI) framework to improve the accuracy of bloodstream infection (BSI) prediction through historical electronic health records (EHRs). Specifically, the paper aims to develop a model that can effectively distinguish between high-risk and low-risk BSI cases, thereby optimizing the allocation of medical resources, reducing unnecessary medical examinations, and ultimately improving patient outcomes. ### Main Issues: 1. **Improving BSI Prediction Accuracy**: Existing clinical decision tools need improvement in predicting BSI and sepsis, particularly in terms of specificity and sensitivity. 2. **Optimizing Medical Resource Allocation**: By more accurately identifying high-risk patients, medical resources can be allocated more effectively, reducing unnecessary examinations for low-risk patients. 3. **Reducing Medical Costs**: Improving prediction accuracy can reduce unnecessary blood cultures and antibiotic use, thereby lowering medical costs. 4. **Early Intervention**: Timely identification of high-risk patients allows for early intervention to prevent disease progression. ### Solution: - **Data Source**: Utilizing electronic health records of 35,591 patients from St. Olav's Hospital in Trondheim, Norway, from 2015 to 2020, including demographic data, laboratory results, and complete medical history. - **Model Development**: Multiple machine learning models were developed, particularly tree-based models (such as XGBoost), which significantly outperformed existing models in terms of specificity and sensitivity metrics. - **Explainability**: An explainable artificial intelligence (XAI) framework was adopted to make the model's predictions more understandable and usable by medical professionals. ### Key Contributions: - **Prediction Accuracy**: The new model demonstrated higher accuracy in predicting BSI, especially in terms of specificity and sensitivity. - **Resource Optimization**: By more accurately identifying high-risk patients, medical resources can be allocated more rationally, reducing unnecessary examinations. - **Clinical Application**: The framework is designed to be easy to understand and use, suitable for medical professionals and users with lower technical expertise, making it a valuable tool in hospital systems. In summary, this paper significantly improves the accuracy of bloodstream infection prediction by developing and validating an explainable artificial intelligence framework based on historical electronic health records, providing strong support for clinical decision-making.

Advancing Bloodstream Infection Prediction Using Explanable Artificial Intelligence Framework

An Artificial Intelligence Approach to Bloodstream Infections Prediction

Predicting bloodstream infection outcome using machine learning

In-hospital mortality, readmission, and prolonged length of stay risk prediction leveraging historical electronic patient records

Using interpretable machine learning to predict bloodstream infection and antimicrobial resistance in patients admitted to ICU: Early alert predictors based on EHR data to guide antimicrobial stewardship

A Machine Learning Predictive Model of Bloodstream Infection in Hospitalized Patients

Moving from Predicting Hospital Deaths by Antibiotic-Resistant Bloodstream Bacteremia Toward Actionable Risk Reduction Using Machine Learning on Electronic Health Records.

Explainable artificial intelligence model to predict acute critical illness from electronic health records

Early diagnosis of bloodstream infections in the intensive care unit using machine-learning algorithms

ICU Bloodstream Infection Prediction: A Transformer-Based Approach for EHR Analysis

Assessment of Sepsis Risk at Admission to the Emergency Department: Clinical Interpretable Prediction Model

Early identification of bloodstream infection in hemodialysis patients by machine learning

Real-Time Artificial Intelligence System for Bacteremia Prediction in Adult Febrile Emergency Department Patients

A Knowledge Distillation Ensemble Framework for Predicting Short and Long-term Hospitalisation Outcomes from Electronic Health Records Data

A New Scoring System to Predict Blood Stream Infections in Patients with Complicated Intra-Abdominal Infections: Experience from a Tertiary Referral Hospital in China

An Interpretable Machine Learning Model for Accurate Prediction of Sepsis in the ICU

Accurate prediction of blood culture outcome in the intensive care unit using long short-term memory neural networks

An Overview of Explainable AI Studies in the Prediction of Sepsis Onset and Sepsis Mortality

Development and Validation of an Interpretable Conformal Predictor to Predict Sepsis Mortality Risk: Retrospective Cohort Study

DeepAISE -- An End-to-End Development and Deployment of a Recurrent Neural Survival Model for Early Prediction of Sepsis

Machine learning-enabled prediction of prolonged length of stay in hospital after surgery for tuberculosis spondylitis patients with unbalanced data: a novel approach using explainable artificial intelligence (XAI)