C-Reactive Protein Testing Does Not Decrease Antibiotic Use for Acute Cough Illness When Compared to a Clinical Algorithm

Abstract

Antibiotics are commonly overused in adults seeking emergency department (ED) care for acute cough illness. To evaluate the effect of a point-of-care C-reactive protein (CRP) blood test on antibiotic treatment of acute cough illness in adults. A randomized controlled trial was conducted in a single urban ED in the United States. The participants were adults (age ≥ 18 years) seeking care for acute cough illness (≤ 21 days duration); 139 participants were enrolled, and 131 completed the ED visit. Between November 2005 and March 2006, study participants had attached to their medical charts a clinical algorithm with recommendations for chest X-ray study or antibiotic treatment. For CRP-tested patients, recommendations were based on the same algorithm plus the CRP level. There was no difference in antibiotic use between CRP-tested and control participants (37% [95% confidence interval (CI) 29-45%] vs. 31% [95% CI 23-39%], respectively; p = 0.46) or chest X-ray use (52% [95% CI 43-61%] vs. 48% [95% CI 39-57%], respectively; p = 0.67). Among CRP-tested participants, those with normal CRP levels received antibiotics much less frequently than those with indeterminate CRP levels (20% [95% CI 7-33%] vs. 50% [95% CI 32-68%], respectively; p = 0.01). Point-of-care CRP testing does not seem to provide any additional value beyond a point-of-care clinical decision support for reducing antibiotic use in adults with acute cough illness.