OBJECTIVE: We derived and tested a protocol to automatically order a chest radiograph (CXR) at emergency department triage for patients with signs and symptoms of pneumonia to reduce time to antibiotics. METHODS: We derived a protocol using a retrospective study of admitted adult patients with pneumonia then prospectively tested the protocol on time to antibiotics. The protocol included patients with a chief complaint of chest pain, shortness of breath, upper respiratory tract infection, hemoptysis, fever, and cough. Of those, patients 50 years or older with any vital sign abnormality and patients younger than 50 with a comorbidity of immunocompromise, cancer, diabetes, transplant, or chronic alcoholism had a CXR ordered automatically. RESULTS: Although the protocol was only 35% (95% confidence interval, 28%-43%) sensitive in identifying patients admitted with pneumonia, time to antibiotics (in hours) (3.4 vs 4.2, P = .01) and time to CXR (3.0 vs 2.0, P = .01) for patients admitted with pneumonia were lower during the study period. CONCLUSION: Automated CXR at triage reduces time to antibiotics in patients admitted with pneumonia.
OBJECTIVE: We derived and tested a protocol to automatically order a chest radiograph (CXR) at emergency department triage for patients with signs and symptoms of pneumonia to reduce time to antibiotics. METHODS: We derived a protocol using a retrospective study of admitted adult patients with pneumonia then prospectively tested the protocol on time to antibiotics. The protocol included patients with a chief complaint of chest pain, shortness of breath, upper respiratory tract infection, hemoptysis, fever, and cough. Of those, patients 50 years or older with any vital sign abnormality and patients younger than 50 with a comorbidity of immunocompromise, cancer, diabetes, transplant, or chronic alcoholism had a CXR ordered automatically. RESULTS: Although the protocol was only 35% (95% confidence interval, 28%-43%) sensitive in identifying patients admitted with pneumonia, time to antibiotics (in hours) (3.4 vs 4.2, P = .01) and time to CXR (3.0 vs 2.0, P = .01) for patients admitted with pneumonia were lower during the study period. CONCLUSION: Automated CXR at triage reduces time to antibiotics in patients admitted with pneumonia.