Predicting the causative pathogen among children with pneumonia using a causal Bayesian network

Background Pneumonia remains a leading cause of hospitalization and death among young children worldwide, and the diagnostic challenge of differentiating bacterial from non-bacterial pneumonia is the main driver of antibiotic use for treating pneumonia in children. Causal Bayesian networks (BNs) serve as powerful tools for this problem as they provide clear maps of probabilistic relationships between variables and produce results in an explainable way by incorporating both domain expert knowledge and numerical data. Methods We used domain expert knowledge and data in combination and iteratively, to construct, parameterise and validate a causal BN to predict causative pathogens for childhood pneumonia. Expert knowledge elicitation occurred through a series of group workshops, surveys and one-on-one meetings involving 6-8 experts from diverse domain areas. The model performance was evaluated based on both quantitative metrics and qualitative expert validation. Sensitivity analyses were conducted to investigate how the target output is influenced by varying key assumptions of particular high degree of uncertainty around data or domain expert knowledge. Results Designed to apply to a cohort of children with X-ray confirmed pneumonia who presented to a tertiary paediatric hospital in Australia, the resulting BN offers explainable and quantitative predictions on a range of variables of interest, including the diagnosis of bacterial pneumonia, detection of respiratory pathogens in the nasopharynx, and the clinical phenotype of a pneumonia episode. Satisfactory numeric performance has been achieved including an area under the receiver-operating characteristics curve of 0.8 in predicting clinically-confirmed bacterial pneumonia with sensitivity 88% and specificity 66% given certain input scenarios (i.e., information that is available and entered into the model) and trade-off preferences (i.e., relative weightings of the consequences of false positive versus false negative predictions). We specifically highlight that a desirable model output threshold for practical use is very dependent upon different input scenarios and trade-off preferences. Three commonly encountered scenarios were presented to demonstrate the potential usefulness of the BN outputs in various clinical pictures. Conclusions To our knowledge, this is the first causal model developed to help determine the causative pathogen for paediatric pneumonia. We have shown how the method works and how it would help decision making on the use of antibiotics, providing insight into how computational model predictions may be translated to actionable decisions in practice. We discussed key next steps including external validation, adaptation and implementation. Our model framework and the methodological approach can be adapted beyond our context to broad respiratory infections and geographical and healthcare settings. Author summary Pneumonia is a leading cause of hospital visits among young children. Doctors need to weigh a range of observations to make timely diagnostic and management decisions, including to differentiate bacterial from viral infections. This is a difficult task to achieve without support due to the complex interactions among relevant factors, and is a major driver of unnecessary antibiotic use. We used domain expert knowledge and data to create a causal Bayesian network (BN) model which depicts an integrated picture of how biological, epidemiological and clinical processes interact to form the challenge of diagnosing and managing children who present to hospital for pneumonia. The model can produce reliable and explainable quantitative inference to help distinguish viral from bacterial infections. We used a few examples to demonstrate how the BN can be used in clinical context, and discussed how computational model predictions may be translated to actionable decisions in practice. We discussed key next steps including the evaluation of the model using external data sets, and adaptation and implementation of the model in other settings. We also discussed how our model framework and the methodological approach can be adapted beyond our context to broad respiratory infections and geographical and healthcare settings.