Predictive model helps reduce need for chest X-rays to diagnose pneumonia in kids
A simple diagnostic model could help pediatric providers reduce the need for chest X-rays to confirm the presence of community acquired pneumonia, according to new research.
Pinpointing this clinical concern typically requires radiographs that may expose children to unnecessary radiation with no additional benefit. Scientists at the Ann & Robert H. Lurie Children’s Hospital of Chicago sought to address this concern, sharing their work Dec. 1 in Pediatrics.
“Our predictive model for community acquired pneumonia is a critical step toward safely reducing radiation exposure in children,” lead author Sriram Ramgopal, MD, an emergency medicine physician and assistant professor of pediatrics at Northwestern University, said in a Monday announcement. “For patients who are determined to be at low risk for pneumonia, we can also avoid unnecessary antibiotic use.”
Ramgopal et al. built their model using X-rays, symptoms, clinical history and other info from more than 1,100 pediatric patients evaluated for pneumonia. Increasing age, fever duration and decreased breath sounds during a stethoscope exam were found to be the three biggest predictors of high- or low-risk pneumonia. The model logged an area under the receiver operating characteristic of 0.80. Out of 229 patients with a predicted risk of less than 4%, 13 (or 6%) had community-acquired pneumonia (for a sensitivity of 95%). On the other side, of those with a predicted risk of greater than 39%, 140 (61%) had pneumonia, for a specificity of 90%.
“A predictive model including age, fever duration, and decreased breath sounds has excellent discrimination for radiographic [community acquired pneumonia],” the authors concluded. “After external validation, this model may facilitate decisions around [chest X-ray] or antibiotic use in CAP.”
Read more about their work in Pediatrics here (payment required).