P2.160: Modified Pneumonia Risk Score Using Lung Ultrasound for the Prediction of Radiographic Pneumonia in Children

Background: Pneumonia is a leading cause of morbidity and mortality in children worldwide, and accurate diagnosis based on clinical symptoms remains a significant challenge. The Pneumonia Risk Score (PRS) is a clinical tool that predicts the risk of radiographic pneumonia in children, with high-risk patients often treated without a chest x-ray (CXR), while moderate-risk patients may still require one. We hypothesized that incorporating lung ultrasound (LUS) findings into the PRS could improve its diagnostic accuracy. Additionally, we evaluated whether the modified PRS using LUS (PRS+LUS) may decrease the need for CXRs.

Methods: We enrolled a prospective convenience sample of children aged 6 months to 18 years presenting to an emergency department with acute respiratory complaints who underwent CXR for suspected pneumonia and LUS for research purposes. The primary outcome of radiographic pneumonia was used as the dependent variable in the logistic regression model with the five predictors from the PRS as independent variables: fever, triage oxygen saturation, age, presence of rales, and presence of wheeze. We calculated the area under the curve (AUC) of this model and compared it to the AUC of the PRS+LUS with consolidation > 1 cm with bronchograms as an additional independent variable. The distribution of patients categorized as low, moderate, and high-risk were compared between models. Finally, we calculated the number of CXRs that could be prevented using the modified PRS+LUS tool.

Results: Among 275 children included in the study, 58 (21.1%) had radiographic pneumonia. The sensitivity of lung consolidation > 1 cm with bronchograms to detect pneumonia was 67.2% (95% confidence interval [CI]: 53.7%, 79.0%), and the specificity was 94.0% (95% CI: 90.0%, 96.8%). The AUC of the PRS was 0.704 (95% CI: 0.633, 0.776), while the AUC of the PRS+LUS was 0.848 (95% CI: 0.786, 0.910; p< 0.001). Furthermore, the PRS+LUS recategorized 40 children (22.5%) from the moderate risk category into the higher risk category, for whom CXR could be prevented.

Conclusion: Adding LUS to the PRS improves the predictive ability for radiographic pneumonia by improving the specificity and sensitivity of the clinical prediction tool. Additionally, the modified PRS+LUS may decrease the need for chest radiographs for patients who have a moderate risk of pneumonia based on clinical signs and symptoms.