Predicting sepsis treatment decisions in the paediatric emergency department using machine learning: the AiSEPTRON
Sylvester Gomes1, Harpreet Dhanoa2, Phil Assheton2
1Evelina London Children's Hospital, London, UK sylvester.gomes@gstt.nhs.uk.
Insights
Machine learning accurately predicts antibiotic use in children with suspected sepsis in emergency departments. Models show moderate accuracy for predicting critical care and serious infections, aiding early intervention.
Area of Science:
- Pediatric Emergency Medicine
- Clinical Informatics
- Artificial Intelligence in Healthcare
Background:
- Early sepsis identification in children is critical for effective treatment and improved outcomes.
- Current pediatric sepsis risk scores are often inadequate for rapid emergency department diagnosis.
- Developing advanced diagnostic tools is essential for timely sepsis management.
Purpose of the Study:
- To create and assess machine learning (ML) models for predicting clinical interventions and patient outcomes in pediatric sepsis cases.
- To enhance the early detection of sepsis in children within emergency department settings.
- To leverage ML for improved sepsis risk stratification in pediatric patients.
Main Methods:
- A retrospective observational study was conducted using electronic health records from a UK tertiary care hospital.
- Machine learning models, including XGBoost, were developed and validated using 15 key predictors from triage and post-blood test data.
- Natural Language Processing (NLP) was employed to integrate unstructured triage note information into the prediction models.
Main Results:
- The triage model predicted antibiotic administration with an AUC of 0.80.
- Models demonstrated moderate accuracy in predicting critical care (AUC 0.78) and serious infection (AUC 0.76).
- Key predictors identified included triage category, temperature, capillary refill time, and C-reactive protein.
Conclusions:
- Machine learning models show significant accuracy in predicting early antibiotic use in pediatric sepsis.
- The models offer moderate predictive power for critical care and serious infection outcomes.
- Further research and external validation are necessary to optimize these ML tools for clinical practice.
Background:
Early identification of children at risk of sepsis in emergency departments (EDs) is crucial for timely treatment and improved outcomes. Existing risk scores and criteria for paediatric sepsis are not well-suited for early diagnosis in ED.
Objective:
To develop and evaluate machine learning models to predict clinical interventions and patient outcomes in children with suspected sepsis.
Design:
Retrospective observational study.
Setting:
ED of a tertiary care hospital, UK.
Patients:
Electronic health records of children <16 years of age attending between 1 January 2018 and 31 December 2019. Patients presenting with minor injuries were excluded.
Methods:
Prediction models were developed and validated, using 15 key predictors from triage and post-blood test data. XGBoost, the best-performing machine learning model, integrated these predictors with triage note information extracted via Natural Language Processing.
Main Outcomes:
(1) Administration of antibiotics; (2) critical care: antibiotics with fluid resuscitation above 20 mL/kg or non-elective mechanical ventilation; (3) serious infection: hospital admission for antibiotics >48 hours.Model performance was evaluated using area under the receiver operating characteristic curve (AUC), likelihood ratios and positive and negative predictive values.
Results:
Triage model: predicted antibiotics at triage (n=35 795; 3.2% with outcome) with an AUC of 0.80 (95% CI 0.76 to 0.84).Antibiotic model: predicted antibiotics post-blood tests (n=4700; 24.2%) with an AUC of 0.78 (95% CI 0.73 to 0.81).Critical care model: predicted critical care (n=4700; 3.3%) with an AUC of 0.78 (95% CI 0.72 to 084).Serious infection model: predicted serious infection (n=4700; 9.4%) with an AUC of 0.76 (95% CI 0.71 to 0.81).Key predictors included triage category, temperature, capillary refill time and C reactive protein.
Conclusion:
Machine learning models demonstrated good accuracy in predicting antibiotic use following triage and moderate accuracy for critical care and serious infection. Further development and external validation are ongoing.
