CLINICAL DECISION RULES TO PEDIATRIC PNEUMONIA

Children'S Hospital Med Ctr (Cincinnati), Cincinnati OH

Investigators

Linked publications & trials

Abstract

Radiographic examination of the chest is done on patients with symptoms of a lower respiratory tract infection (LRI) primarily to confirm or exclude the presence of a radiographic infiltrate consistent with pneumonia. However, the 1995 National Health Survey showed that the prevalence of pneumonia in children under 5 years of age with respiratory infection was only approximately 2.8%. Even among pediatric patients presenting to the emergency department (ED), radiographic pneumonia was present in only 7% to 19% of patients with LRI symptoms. Despite this relatively low proportion of chest radiographs (CXRs) that reveal radiographic pneumonia, many such patients are routinely referred for a CXR by their primary care or ED physician. As a consequence, patients with respiratory infection, but without radiographic pneumonia, often have CXRs, resulting in excess radiation exposure and additional cost. Numerous studies have been published on pediatric pneumonia, however there is still no consensus opinion about how to accurately diagnose this disease based on a combination of historical and physical examination findings, nor is there agreement that a CXR is always needed to diagnose this disease in children with respiratory symptoms. The purpose of this study is to develop decision rules that will predict radiographic pneumonia in patients age 2 months to 59 months of age with symptoms of LRI. Subsequent studies will validate the decision rules and determine the cost effectiveness of applying the rules. This study will be conducted in the ED of Children's Hospital Medical Center during October 2000 through September 2001. All patients with LRI symptoms will be approached for enrollment. Detailed history and physical examination variables will be assessed on each patient and all patients will receive a CXR. Each CXR will be reviewed by 2 blinded radiologists and a consensus reading will be obtained. Variables that are strongly associated with radiographic pneumonia (p<0.05) will be analyzed by multivariate techniques. Decision rules will be developed of the best combination of predictor variables for pneumonia on CXR.