Original Article: Identifying Outpatients with Acute Cough at Very Low Risk of Pneumonia [Point-of-Care Guides]

Issue Date: August 15, 2019

See additional reader comments at: https://www.aafp.org/afp/2019/0815/p246.html

To the Editor: Given how common it is for patients to present with acute cough in the primary care setting, I read Dr. Ebell’s article with great interest, and I agree with his assessment of the trials mentioned. However, I would point out that the authors of the GRACE study, as well as the two similar U.S. studies mentioned, make the assumption that acute cough in the setting of positive chest radiograph findings establishes a diagnosis of community-acquired pneumonia (CAP) and warrants antibiotic use. Although many experts would agree with this assumption, others require the patient to meet more rigorous clinical and laboratory criteria before diagnosing CAP and prescribing antibiotics.¹

This lack of consensus stems from a dearth of data on which patients with acute cough benefit from antibiotics and which do not. Indeed, only two placebo-controlled trials exist for patients with CAP.^2,3 These trials showed the benefit of antibiotics, but they also used more stringent inclusion criteria than simply the presence of acute cough and suggestive chest radiograph findings—both required patients to have a fever, and one required patients to have confirmed Mycoplasma pneumoniae infection.^2,3

There may be a subset of patients with acute cough and suggestive radiograph findings who do not benefit from antibiotics, and even a subset of patients with a particular constellation of symptoms and negative chest radiograph findings who do benefit. We simply do not have adequate evidence to definitively state which patients will and will not benefit from antibiotics. A far more helpful study to reduce unwarranted antibiotic use would determine which patients with acute cough benefit from antibiotics and which do not, instead of which patients with acute cough will likely have positive chest radiograph findings, as existing studies demonstrate.

Finally, any discussion of identifying a patient’s risk of CAP should include the increasingly widespread use of bedside ultrasonography. A lung examination with ultrasonography takes less than one minute to perform and has superior accuracy to chest radiography when using chest computed tomography as the reference standard.⁴ However, it is unclear if a patient with positive ultrasound findings, especially in the context of negative radiograph findings, should receive antibiotics, underscoring the need for studies evaluating which patients with acute cough benefit from antibiotics.

In Reply: I thank Dr. Tanael for his thoughtful letter and agree with his comments, with some caveats. He correctly notes that not all patients with radiographic CAP benefit from an antibiotic. However, because approximately 70% of patients with acute cough receive an antibiotic,¹ and only 4% of primary care patients with cough are diagnosed with CAP,² the larger task is reducing inappropriate antibiotic use among those without CAP rather than in those with CAP. To that end, identifying patients who are unlikely to have radiographic CAP may be helpful. In addition, as he notes, data are lacking regarding which patients with radiographic CAP benefit from an antibiotic. However, a study found that C-reactive protein has independent predictive value for identifying lower respiratory tract infections caused by a bacterial pathogen.³ My colleagues and I are in the process of gathering prospective data on 1,400 patients with acute cough, to learn more about how to identify patients with acute cough who are unlikely to benefit from antibiotics. This study is funded by the Agency for Healthcare Research and Quality and involves data collection in Madison, Wis.; Washington, D.C.; and Athens, Ga. Reducing inappropriate antibiotic use in patients with nonpneumonia lower respiratory tract infections by 30% would yield a much larger benefit than reducing antibiotic use in patients with CAP by the same amount.

Regarding ultrasonography, I agree that it has good accuracy for the diagnosis of pneumonia in the hands of adequately trained clinicians. However, as Dr. Tanael notes, it is unclear how ultrasound-diagnosed CAP differs from radiographically diagnosed CAP in terms of the benefit (or lack of benefit) of antibiotics.