Models Predict Severity of Pneumonia in Kids to Help Guide Treatment

Community-acquired pneumonia is a leading cause of infection in children globally and is one of the most common and costly reasons for pediatric hospitalization in the United States. While most children with pneumonia recover fully after a mild illness, about 5 percent develop severe symptoms and complications. Although severe outcomes are rare, identifying these high-risk patients early is crucial for clinicians to take swift action and prevent further deterioration. It is equally important to identify cases that are likely to remain mild, so as to avoid unnecessary tests, treatments, or hospital stays. Now, researchers have developed practical models that effectively distinguish between mild, moderate, and severe pneumonia in children, based on a study conducted across 73 Emergency Departments (EDs) in 14 countries, through the international Pediatric Emergency Research Network (PERN). These new predictive tools are designed to assist clinicians in determining whether a child’s pneumonia requires hospitalization or intensive care.

The study, led by researchers at Lurie Children’s Hospital of Chicago (Chicago, IL, USA), involved over 2,200 children aged 3 months to 14 years, all of whom presented with community-acquired pneumonia in the ED. The researchers found that children with pneumonia who had symptoms such as a runny nose and congestion were more likely to have a mild illness. They also identified clinical indicators associated with moderate or severe pneumonia, for which hospitalization should be considered. These signs include abdominal pain, refusal to drink fluids, being on antibiotics for the current illness before the ED visit, chest retractions (which suggest difficulty breathing), respiratory or heart rates above the 95th percentile for age, and hypoxemia (low oxygen levels in the blood).

These clinical features are typically assessed in patients with respiratory illnesses, making the model applicable and easy to implement in clinical practice. For emergency departments worldwide, which see thousands of pediatric pneumonia cases daily, there has previously been no reliable method for predicting which children are truly at risk of worsening. This new model offers clinicians a practical, data-driven tool to aid decision-making, ultimately enhancing care and improving outcomes. The study, published in the journal Lancet Child and Adolescent Health, also included models to predict pneumonia severity specifically in children with pneumonia visible on chest radiographs. In addition to the previously mentioned clinical features, the researchers found that the risk of severe illness increased if multiple areas of the lung were affected.

“Our pediatric pneumonia predictive models show good-to-excellent accuracy,” said lead author Todd Florin, MD, MSCE. “They appear to perform better than clinician judgment alone in predicting illness severity, according to previous research from Lurie Children’s. Once externally validated, our models will provide evidence-based information for clinicians to consider when evaluating pneumonia in children.”