Related Experiment Video
Updated: Jun 28, 2025
Prediction of Red Blood Cell Antibody Significance Using the Monocyte-Macrophage Assay
Published on: February 7, 2025
Predicting community acquired bloodstream infection in infants using full blood count parameters and C-reactive
Lieke Brouwer1,2, Robert Cunney3,4, Richard J Drew3,4,5
1Public Health Laboratory, HSE, Cherry Orchard Hospital, Dublin, Ireland. lieke.brouwer@hse.ie.
Insights
Machine learning models can identify infants unlikely to have bloodstream infections (BSI) using blood count and CRP tests. This can help avoid unnecessary antibiotic treatments while awaiting blood culture results.
Area of Science:
- Pediatric Infectious Diseases
- Clinical Machine Learning
- Biomarker Discovery
Background:
- Bloodstream infections (BSI) in infants present with non-specific symptoms, complicating early diagnosis.
- Blood culture results, the gold standard for BSI diagnosis, require up to 48 hours, often leading to empirical antibiotic use.
- Accurate and timely identification of infants without BSI is crucial to prevent unnecessary antibiotic exposure and diagnostics.
Purpose of the Study:
- To develop and evaluate predictive models for identifying infants unlikely to have BSI.
- To utilize routinely available clinical data, specifically Full Blood Count (FBC) and C-reactive protein (CRP) levels.
- To reduce the burden of unnecessary antibiotic treatments in infants with suspected BSI.
Main Methods:
- Trained multiple machine learning models (logistic regression, LDA, kNN, SVM, random forest, decision tree) on data from 2693 infants (7-60 days old) with suspected BSI.
- Utilized FBC and CRP values from infants treated between 2005 and 2022 at a tertiary pediatric hospital.
- Validated the best performing models (decision tree, random forest) on the full dataset and a separate 2023 dataset.
Main Results:
- All tested models demonstrated comparable sensitivities (47%-62%) and specificities (85%-95%).
- Decision tree and random forest models effectively stratified infants into low- and high-risk groups for BSI.
- Negative predictive values were high (> 99% for full dataset, > 97% for 2023 dataset), indicating high confidence in ruling out BSI in low-risk infants.
Conclusions:
- Developed machine learning models capable of predicting negative blood cultures in infants aged 7-60 days with suspected BSI.
- These models show potential to guide clinical decisions, reducing unnecessary antibiotic administration and diagnostic procedures.
- Implementation of these models can improve antibiotic stewardship and patient outcomes in pediatric care.
Abstract:
Early recognition of bloodstream infection (BSI) in infants can be difficult, as symptoms may be non-specific, and culture can take up to 48 h. As a result, many infants receive unneeded antibiotic treatment while awaiting the culture results. In this study, we aimed to develop a model that can reliably identify infants who do not have positive blood cultures (and, by extension, BSI) based on the full blood count (FBC) and C-reactive protein (CRP) values. Several models (i.e. multivariable logistic regression, linear discriminant analysis, K nearest neighbors, support vector machine, random forest model and decision tree) were trained using FBC and CRP values of 2693 infants aged 7 to 60 days with suspected BSI between 2005 and 2022 in a tertiary paediatric hospital in Dublin, Ireland. All models tested showed similar sensitivities (range 47% - 62%) and specificities (range 85%-95%). A trained decision tree and random forest model were applied to the full dataset and to a dataset containing infants with suspected BSI in 2023 and showed good segregation of a low-risk and high-risk group. Negative predictive values for these two models were high for the full dataset (> 99%) and for the 2023 dataset (> 97%), while positive predictive values were low in both dataset (4%-20%). Conclusion: We identified several models that can predict positive blood cultures in infants with suspected BSI aged 7 to 60 days. Application of these models could prevent administration of antimicrobial treatment and burdensome diagnostics in infants who do not need them. What is Known: • Bloodstream infection (BSI) in infants cause non-specific symptoms and may be difficult to diagnose. • Results of blood cultures can take up to 48 hours. What is New: • Machine learning models can contribute to clinical decision making on BSI in infants while blood culture results are not yet known.
Related Concept Videos
01:18Steps in Outbreak Investigation
01:30Pneumonia III: Complications and Assessment