Pediatric Pneumonia Without Routine Chest X-Rays: How Much Weight Should We Give the Lung Exam?

Have you stood at the bedside while the resident, nurse, and student each described a different lung examination? Someone heard focal crackles. Someone else heard coarse breath sounds. The next clinician heard clear lungs. The PECARN PedCAPS (Pediatric Community-Acquired Pneumonia Severity) study suggests that this familiar disagreement is common.

In a new PECARN PedCAPS analysis published in JAMA Network Open, two clinicians independently examined the same 252 children with community-acquired pneumonia (CAP). No individual physical examination finding met the study’s prespecified reliability standard [1]. A single lung sound should not decide the diagnosis of pneumonia.

This creates a clinical dilemma.

The Pediatric Infectious Diseases Society (PIDS)/Infectious Diseases Society of America (IDSA) guidelines advise against routine chest radiography to confirm suspected CAP in children well enough for outpatient treatment [2]. However, the clinician must still decide whether CAP is likely enough to treat. That decision depends heavily on the history and physical examination, even though no individual symptom or sign is sufficiently sensitive or specific and clinicians may interpret the same lung sounds differently [3].

How do we then diagnose outpatient CAP without reflexively ordering an x-ray?

Study Design

This planned reliability analysis was nested within the ongoing prospective PECARN PedCAPS study cohort. Investigators enrolled children ages 3 months through 17 years with clinician-diagnosed CAP at seven academic pediatric EDs. Two attending physicians, advanced practice clinicians, or PEM fellows independently examined each child within 60 minutes. The clinicians completed identical standardized forms but received no special training in how to define the examination findings.

The study population limits how far we can take the results. Nearly all participants, 249 of 252 (98.8%), had pneumonia supported by chest radiography, and 128 (50.8%) were hospitalized. This selected group had established CAP and differs from the undifferentiated population of children presenting with cough, fever, or respiratory distress.

Investigators measured raw agreement and Fleiss κ for the following:

• Altered mental status
• Capillary refill
• Grunting
• Retractions or accessory muscle use
• Auscultatory lung findings:
  • Wheezing
  • Decreased breath sounds
  • Crackles or rales
  • Rhonchi or coarse breath sounds

Acceptable reliability required the lower bound of the κ 95% confidence interval to be at least 0.40.

Results

No finding passed the prespecified threshold. Wheezing and retractions performed best, but only reached moderate agreement:

• Wheezing: κ = 0.50 (95% CI 0.39-0.62), with 82% raw agreement
• Retractions or accessory muscle use: κ = 0.49 (95% CI 0.37-0.60), with 78% raw agreement
• Crackles or rales: κ = 0.21 (95% CI 0.11-0.31), with 55% raw agreement
• Rhonchi or coarse breath sounds: κ = 0.18 (95% CI 0.08-0.28), with 56% raw agreement
• Decreased breath sounds: κ = 0.11 (95% CI 0.02-0.20), with 46% raw agreement

When wheezing was simplified to present or absent without asking clinicians to distinguish focal from diffuse findings, κ improved to 0.58. This suggests that localization added another layer of disagreement.

Reliability did not differ significantly between the 124 children discharged home and the 128 hospitalized. The analyses by age, race, and clinician knowledge of chest radiograph results were underpowered and should not be described as showing equivalence across those groups.

Altered mental status, delayed capillary refill, and grunting were uncommon. Their high overall raw agreement largely reflected agreement that the findings were absent. When at least one examiner documented one of these findings, agreement that it was present was low. Clinicians should still treat altered mental status and grunting as concerning when present.

How to Diagnose Without a Routine Chest X-Ray?

Reliability and diagnostic accuracy are different questions. This study shows that clinicians frequently disagree about examination findings in children already diagnosed with CAP. It did not measure the sensitivity, specificity, or likelihood ratios of those findings among all children with possible pneumonia. Because almost every participant had radiographic pneumonia, it also cannot determine how auscultation performs in the less selected population we evaluate on shift.

The PIDS/IDSA guideline discourages routine imaging while preserving selective imaging when uncertainty matters. It does, however, recommend a chest radiograph when a child has suspected or documented hypoxemia, significant respiratory distress, or failure of initial antibiotic therapy [2]. In a well-appearing child without those features, a clinician should holistically combine the history with measured respiratory rate, oxygen saturation, work of breathing, overall appearance, serial examinations, and a reliable follow-up plan. One reported crackle, focal decrease in breath sounds, or coarse sound should not become the diagnostic push for radiographs.

“Our results emphasize the need to look at multiple features of the clinical presentation rather than exam findings in isolation when making the diagnosis of pneumonia.”

Todd A. Florin, MD, MSCE, senior author of this 2026 JAMA publication

Bottom Line

Keep listening to the lungs, but do not let one auscultatory finding decide whether a child has pneumonia. When examinations conflict and the uncertainty would change management, return to the full clinical picture and obtain additional information selectively.

References

1. Hooli S, Reeder R, Cutler L, et al. Reliability of Physical Examination Findings in Youths Diagnosed With Pneumonia. JAMA Netw Open. 2026;9(8):e2630074. PMID: 42623068. doi:10.1001/jamanetworkopen.2026.30074
2. Bradley JS, Byington CL, Shah SS, et al. The Management of Community-Acquired Pneumonia in Infants and Children Older Than 3 Months of Age: Clinical Practice Guidelines by the Pediatric Infectious Diseases Society and the Infectious Diseases Society of America. Clin Infect Dis. 2011;53(7):e25-e76. PMID: 21880587. doi:10.1093/cid/cir531
3. Lipsett SC, Neuman MI. Rethinking Auscultation for Pediatric Pneumonia: Do My Ears Deceive Me? JAMA Netw Open. 2026;9(8):e2630083. PMID: 42623079. doi:10.1001/jamanetworkopen.2026.30083