The Negative Predictive Value of a CXR in the Evaluation of Pediatric Pneumonia

William T. Basco, Jr., MD, MS


November 20, 2018

The Predictive Value of a CXR in Suspected Pneumonia

Clinicians often face uncertainty about whether they can rely on a negative chest x-ray (CXR) to rule out pneumonia in a child, even when other clinical signs and symptoms may have raised that concern.[1] In a recent study, Lipsett and colleagues[1] aimed to determine the negative predictive value of a CXR in children with suspected pneumonia.

For this prospective observational cohort study, data were collected over a 2-year period in a single large pediatric emergency department (ED). The patients were children and adolescents, aged 3 months to 18 years, who underwent a CXR to evaluate for pneumonia. The general standard of care in this ED was to obtain a CXR when any child was suspected clinically of having pneumonia. Children with chronic conditions that may predispose them to pneumonia were excluded, as were any children already on antibiotic treatment.

CXR findings were classified as "positive," "negative," or "equivocal," depending on the final reading by the radiologist. The investigators followed up with the families by phone or email during the first and second weeks after the ED visit to determine such outcomes as symptom resolution, whether the child had returned to school or day care, treatments rendered, and any additional medical visits required. They specifically asked whether the child had been diagnosed with pneumonia after the index ED visit. For analysis purposes, a child was considered to have pneumonia if the ED discharge diagnosis was pneumonia, regardless of CXR findings.

The study cohort included 683 children with a median age of 3.1 years. More than 20% of the children were hospitalized in conjunction with the index ED visit and 29.3% were diagnosed with pneumonia in the ED. Among the 411 children with a negative CXR who were not diagnosed with pneumonia and not treated with antibiotics, five children (1.2%) were diagnosed with pneumonia within 2 weeks of the index ED visit. Therefore, the negative predictive value of the CXR for pneumonia was 98.8% (95% confidence interval, 97.0% to 99.6%). All five children subsequently diagnosed with pneumonia were < 3 years old and were early in the course of their illnesses when they presented to the ED.

The conclusion of the study was that a negative CXR excludes pneumonia in the large majority of children evaluated for pneumonia. These children can be observed without antibiotic therapy after an ED evaluation and a CXR.


The framework from which to view this study is that pneumonia remains a clinical diagnosis. Some children in the study were treated for pneumonia even with negative CXRs, and others with positive CXRs were not treated. Those may be valid decisions and pose questions not examined in this study.

That said, this study does provide valuable information for what is a fairly narrow question: If a clinician is concerned about pneumonia in a child who does not seem to need treatment based on a history and physical, can a CXR help make the decision about whether to prescribe antibiotics? The answer would appear to be yes. Those children with negative CXRs who have not already crossed a clinician's threshold for requiring antibiotics are very unlikely to ever cross that threshold. I can't emphasize enough that the clinician's decision-making is important: clinicians should use CXR findings to supplement their decisions and not to drive them entirely.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.