Objective: National guidelines discourage routine chest radiographs (CXRs) to confirm suspected pneumonia in children managed as outpatients. However, limiting CXRs may lead to antibiotic overuse. We examined the impact of CXRs and clinical suspicion on antibiotic treatment for children with suspected pneumonia.
Methods: Children aged 3 months to 18 years undergoing CXR for suspected pneumonia in a pediatric emergency department were prospectively enrolled. Before CXR, physicians indicated their initial plan for antibiotics (yes or no) and clinical suspicion for radiographic pneumonia (<5%, 5-10%, 11-20%, 21-50%, 51-75%, >75%). Subjects had radiographic pneumonia if their CXRs demonstrated definite or possible findings of pneumonia. We compared antibiotic treatment according to pre-CXR antibiotic plan and suspicion for pneumonia and CXR results.
Results: Among the 107 children with a plan for antibiotics before CXR, 72% ultimately received antibiotics compared with 19% of the 1503 children without a pre-CXR plan for antibiotics (P < 0.001). Among those patients with a pre-CXR plan for antibiotics, 96% of children with radiographic pneumonia were ultimately treated compared with 54% without radiographic pneumonia (P < 0.001). If antibiotics were not initially planned, 37% with radiographic pneumonia were treated compared with 8% without radiographic pneumonia (P < 0.001). The use of CXR was more likely to influence antibiotic prescribing patterns when the clinical suspicion of pneumonia was low (<20%).
Conclusions: Among children with high suspicion for pneumonia, CXRs infrequently altered the initial plan for antibiotics. However, when clinical suspicion for pneumonia was low, the use of CXR may reduce unnecessary antibiotic use.