Pneumonia with a Negative Chest Film: Treat with Antibiotics or Simply Observe?
LewisFirst, MD, MS, Editor in Chief, Pediatrics
How often do you suspect pneumonia clinically by history and physical exam only to
find that the chest radiograph (CXR) is negative? When this happens, how do you decide
whether or not to treat with an antibiotic? Lipsett et al. (10.1542/peds.2018-0236) decided to examine the negative predictive value of a CXR in an interesting study
being released this week in our journal. The authors enrolled 683 children in a two-year
study who underwent a CXR for suspected pneumonia in an emergency department. Those
already on antibiotics or with an underlying chronic disease were excluded. Follow-up
included a two-week period to make sure that complications of not treating or missing
a bacterial pneumonia did not occur. Of the 457 children with negative films, 411
were not treated with an antibiotic and only 5 of these children were diagnosed with
pneumonia over the next two weeks, resulting in a negative predictive value of 98.8%.
The authors conclude that in the absence of high clinical suspicion for pneumonia,
antibiotics should not be used if the CXR is negative. So how convincing is this
study to enable you to be a better steward of antibiotics when you detect clinical
signs of pneumonia, but the chest film is negative?
To better answer that question, we asked resource conscious hospitalists Dr. Matthew
Garber and Ricardo Quinonez to weigh in with an accompanying commentary (10.1542/peds.2018-2025). Drs. Garber and Quinonez point out that before we can hold back on using antibiotics,
perhaps an alternative question that needs answering is what is the positive predictive
value of a CXR for diagnosing a community acquired pneumonia? The authors note that
other studies suggest that value is much lower than the negative predictive value
(about 15% of positive films suggest a bacterial pneumonia) and thus argue that we
consider holding antibiotics for even positive radiographs of pneumonia if the clinical
assessment is non-worrisome—meaning maybe we shouldn’t be getting CXRs as much as
we tend to do so—especially given how many positive films represent viral etiologies
rather than bacterial ones based on studies cited in this important commentary. Both
the study and commentary are well worth choosing wisely to read if you find yourself
wondering if your patient with cough, some rales at the right base, and fever but
stable vital signs requires antibiotics and maybe not even a chest film at all. Read
both articles about the value of CXRs and see what develops.