Purpose: To develop a prediction rule for the use of chest radiographs in evaluating for community-acquired pneumonia (CAP) based on presenting signs and symptoms.
Patients and methods: Adult patients with acute respiratory symptoms and positive chest radiographic results from October 2004 through April 2005 were enrolled as positive cases (n = 350). An equal number of age-matched controls with acute respiratory symptoms but negative radiographic results were included. Data analyses were performed on the 6 most common individual clinical indicators (cough, sputum production, fever, tachycardia, tachypnea, and abnormal physical examination results). Additional analyses were performed for any vital sign abnormality and for the presence of vital sign or physical examination abnormalities.
Results: The data show that vital sign and physical examination findings are useful screening parameters for CAP, demonstrating a sensitivity of 95%, a specificity of 56%, and an odds ratio of 24.9 [corrected] in the presence of vital sign or physical examination abnormalities. In light of these results, the authors developed a prediction rule for low-risk patients with reliable follow-up, which states that chest radiographs are unnecessary in the presence of normal vital signs and physical examination findings.
Conclusion: The data suggest that chest radiographs are unnecessary in patients with acute respiratory symptoms who present with normal vital signs and physical examination findings. Because approximately 5% of cases would be missed, however, these criteria are useful only for patients with reliable follow-up and a low likelihood of morbidity if CAP is not detected initially.