Patients with acute bronchitis present primarily with cough and other mild symptoms such as nasal congestion and malaise. Since most cases of acute bronchitis are of viral etiology, antibiotics are not recommended.1 In fact, several systematic reviews and meta-analyses conclude that the risk of adverse effects from antibiotics outweighs the negligible decrease in symptom duration observed in some trials.2-5 Despite these evidence-based recommendations, antibiotic use for acute cough remains high in some areas.6 Perhaps antibiotics are more likely to be given to patients with cough and other more severe symptoms suggestive of pneumonia, a lower respiratory tract infection where antibiotics are indicated.
To examine if symptom severity influences antibiotic use in adults presenting with acute cough and whether antibiotics hasten recovery from acute bronchitis, Butler and colleagues published the results of a prospective, cross sectional observational study.7 Investigators recruited patients from 14 primary care research networks in 13 European countries. Clinicians recorded initial symptoms and subsequent management (antibiotic choice and treatment duration) using a standardized form. Patients documented daily symptoms in a diary for up to 28 days after presentation. Primary outcome measures included antibiotic prescription and patient recovery rates after controlling for initial symptom severity scores.
Primary care providers completed case report forms for 3,368 patients. Patients most commonly complained of cough (99.7%, 3358) and feeling generally unwell (80.1%, 2698). Only 12.4% of patients had an oral temperature >99°F. Clinicians prescribed a median 7-day course of antibiotics for 53% of patients overall, although antibiotic use ranged from 20.6% in the Barcelona network to 87.6% in the Bratislava network. Symptom severity on initial presentation did not correlate with antibiotic prescribing patterns. Amoxicillin, macrolides/lincosamides, amoxicillin-clavulanate and tetracyclines represented 29%, 26%, 15% and 14% of all prescriptions, respectively. Antibiotic selection differed greatly across networks with fluoroquinolones accounting for 18% of prescriptions in the Milan network. Although antibiotic choice may have been guided by local antibiotic susceptibility data, this information was not provided. After accounting for initial clinical presentation, antibiotics decreased symptom severity scores by 0.1% after seven days (p<0.05). The investigators considered this to be clinically unimportant and consistent with a placebo effect.
Clinicians in this study relied heavily on reported symptoms to determine the cause of acute cough. Although most presented with mild symptoms consistent with acute bronchitis rather than pneumonia, providers prescribed antibiotics to more than half of the study patients. Physical exam findings, results from chest radiography, white blood cell counts and pulse oximetry readings were not collected. It is possible that prescribers may have been more likely to withhold antibiotics if this supportive information was used to further exclude pneumonia as a cause of acute cough. Also, since participating clinicians were affiliated with a research network and were aware of the ongoing study, antibiotic prescription estimates reported in this study may underestimate observed habits for general practitioners. While the authors found that antibiotics hastened recovery using an objective measure of symptom severity, the clinical relevance of these findings is unclear since the symptom severity scale has not been previously validated as an appropriate tool for evaluating acute bronchitis. Finally, results from ...