Chest radiograph screening for severe acute respiratory syndrome in the ED

The purpose of the study was to evaluate the use of chest radiography for the screening of severe acute respiratory syndrome (SARS). We retrospectively analyzed all patients who attended an Emergency Department SARS screening clinic during the outbreak in Hong Kong, from March 10 to June 5, 2003. Patients with clinical and epidemiologic suspicion of SARS were evaluated by serial chest radiography. All radiographs were reported by consensus from 2 radiologists, blinded to the clinical records. The prevalence of SARS was 13.3% among 1328 patients included. The initial radiograph had sensitivity 50.3%, specificity 95.0%, positive likelihood ratio 10.06, negative likelihood ratio 0.52, positive predictive value 61.5%, and negative predictive value 92.3% for diagnosing SARS. Serial chest radiography had sensitivity 94.4%, specificity 93.9%, positive likelihood ratio 15.48, negative likelihood ratio 0.06, positive predictive value 71.4%, and negative predictive value 99.0%. The initial chest radiograph has poor sensitivity, and serial radiographs are required to rule out SARS.