Is C-reactive protein testing useful to predict outcome in patients with acute bronchitis?

Abstract

A recent clinical trial could not find differences between anti-inflammatory drugs, antibiotics and placebo in shortening the duration of symptoms in acute bronchitis. To investigate if C-reactive protein (CRP) concentrations at presentation are predictive of symptom resolution in these patients. We performed a secondary analysis of the data from a placebo-controlled, randomized clinical trial carried out in primary care. Patients from 18 to 70 years of age presenting a respiratory tract infection of <1 week of evolution, with cough as the predominant symptom and the presence of discoloured expectoration, were enrolled in the study. On the baseline visit, CRP was determined in capillary blood and a five-item symptom diary was given. Patients were followed up to 30 days. The main outcome measure was the number of days with persistent cough. A total of 312 subjects fulfilled all the criteria for the efficacy analysis and had undergone the CRP test; of these, 56.4% presented a CRP value <8 mg/l and 76% presented <20 mg/l. There were no significant differences in the median duration of cough: 10 days among patients with CRP concentrations <8 mg/l [95% confidence interval (CI): 8-11 days], 11 days among those with concentrations ranging from 8 to 19 mg/l (95% CI: 8-16) and 11 days in those with CRP >20 mg/l (95% CI: 9-12) (P = 0.337). Among patients with uncomplicated acute bronchitis and discoloured sputum, the CRP concentrations at presentation are not helpful for predicting symptom resolution.