• 
• 

Three Part Question

In [infants with bronchiolitis] is [chest xray] a useful investigation [to predict severity or alter management]

Clinical Scenario

A 9 month old infant is brought to the Emergency Department by his mother with a 3 day history of coryzal symptoms and increasing difficulty breathing. Ausculation reveals widespread wheeze and crepitations, and you make a clinical diagnosis of bronchiolitis. You wonder whether a chest xray is indicated to confirm this diagnosis and assess its severity.

Search Strategy

Medline 1950-March 2012 AND embase 1980- March 2012
(exp Bronchiolitis, viral/ OR exp Bronchiolitis/OR bronchiolit*.af OR RSV.af OR respiratory syncytial virus.af) AND (exp x-rays/OR xray.af OR x-ray.af OR radiograph*.af OR exp RADIOGRAPHY/ OR roentgenogram.af OR imaging.af) (Limit to: humans and (age groups all infant birth to 23 months) and English language).

Search Outcome

Medline returned 368 results and Embase returned 2570 results of which eight unique papers directly addressing the BET question were identified.

Relevant Paper(s)

Author, date and country	Patient group	Study type (level of evidence)	Outcomes	Key results	Study Weaknesses
Friis et al 1990 Denmark	128 children aged 1 month to 6 years admitted to hospital with suspected lower respiratory tract infection (bronchiolitis or pneumonia) CXR findings correlated with virology from NPA and bacteriology from nasotracheal secretions	Observational study	x-ray Findings in virus positive (n=76) vs virus negative (n=52) patients irrespective of bacterial findings	'Normal x-ray' 21% vs 8% p<0.05 'Bronchopneumonia' 18% vs. 6% p<0.05 ‘Peribronchitis’: 25% vs 46% p<0.025 No significant difference in rate of lobar pneumonia, hyperinflation, atelectasis, hilar adenopathy.	Includes older children who fall outside the usual age range for bronchiolitis. Difficulties in obtaining secretions from lower airways for bacteriology
			CXR findings in RSV +ve age <6 months vs >6 months	'Lobar' pneumonia more common in <6 month (p<0.025)
			x-ray Findings of in patients with positive versus negative bacteriology	No significant difference
			Lobar pneumonia	Lobar pneumonia: (30%/21%/36%/37% NS),
			x-ray findings in patients with virus+/bact+, virus+/bact-, virus-/bact+& virus-/bact -	bronchopneumonia: (11%/26%/4%/7% NS), interstitial pneumonia: (41%/18%/44%/30% NS) Peribronchitis: (30%/21%/48%/44% NS) Hyperinflation: (24%/15%/20%/11% NS) Atelectasis: (5%/5%/20%/11% NS)
Schuh et al April 2007 Canada	265 infants aged 2-23 months with typical bronchiolitis attending the emergency department All patients had chest x-ray. Study appropriately powered	Prospective Cohort Study. Rate of radiographic alternate diagnosis of patients with typical bronchiolitis	Xray inconsistent with bronchiolitis	2 of 265 (0.75%) 95% CI 0-1.8	Convenience sample- 665 patients not included as attended overnight- may indicated selection bias
			Difference in admission rate pre and post radiography	Same in 258 0f 265 cases (97.4%)
			Correlation of clinical status vs. xray findings	Infants with sats >92% and RDAI score <10 more likely to have a simple radiograph (OR 3.9; 95% CI, 1.3-14.3)
			Rate of antibiotic prescription	Pre-x-ray 7/265 (2%) Post-x-ray 39/265 (14.7%) 95% CI for difference in agreement 0.08 to 0.16
Mahabee-Gittens et al 1999 USA	270 children <18 months of age presenting to paediatric ED with wheeze on physical examination who had CXR performed. Excluded bronchopulmonary dysplasia, congenital heart disease, cystic fibrosis, FB inhalation.	Retrospective chart review	Alternative diagnosis	2/270 (0.7%)	Includes those with previous episodes of wheeze. Clinical data often incomplete. Not all children x-rayed (470 attendees, 270 x-rayed) hence probable selection bias. Reporting radiologist not blinded
			Focal infiltrate on CXR versus clinical findings	History fever: OR 2.1 (p=0.03) T>38.4 in ED: OR 2.5 (p=0.01) Crackles on examination: OR 3.9 (p=0.0002) (after bonferroni correction only crackles significant p<0.005) No significant correlation to Sa02 <93, RR>60, retractions or wheeze
Dawson et al 1990 New Zealand	153 children <6 months of age, admitted to paediatric department with a clinical diagnosis of bronchiolitis	Retrospective study	Radiological severity determined by 2 blinded radiologists versus clinical severity score	No statistically significant relationship between radiological severity grading and clinical severity score	Retrospective study only included children who had x-ray as part of their routine management therefore may have selected those with more severe disease
Eriksson et al, 1986, Norway	137 patients (0–48 months) with proven RSV infection underwent supine CXR and nasopharyngeal swab for bacteriology	Observational study	x-ray Changes in patients with positive vs negative nasopharyngeal swabs	14 vs 4 p=0.01	Inclusion criteria not clearly defined. 14 children clinically had URTI. Included children outside the age range from bronchiolitis. Clinical significance of positive nasopharyngeal bacteriology uncertain
			Atelectasis	No significant difference
			Infiltrates	No significant difference
			Hyperinflation	No significant difference
			Hilar gland enlargement	No significant difference
Yong et al, 2009 Canada	265 previously healthy infants 2-23 months, presenting to an urban tertiary ED with ‘typical bronchiolitis’. Pre-radiograph and post-radiograph ED diagnosis correlated with radiology report from blinded radiologist	Prospective observational study and cost effectiveness analysis	Rate of alternative diagnosis	2/265 (0.8%)	No attempt to correlate x-ray findings with microbiological results or disease severity. ‘Bronchiolitis associated pneumonia’ on x-ray unlikely to represent bacterial pneumonia in this cohort. Cost effectiveness analysis not generalisable to other healthcare systems
			Sensitivity of ED physician for detection alternative diagnosis pre vs.post x-ray	0% vs 0% 96% vs 88%
			Specificity of ED physician for detection alternative diagnosis pre and post x-ray	Pre-x-ray 12% Post x-ray 41%
			Sensitivity of ED physician diagnosis of ‘bronchiolitis associated pneumonia’	Pre-x-ray 90% Post x-ray 83%
			Specificity for ‘bronchiolitis associated pneumonia’	Pre-x-ray 90% Post x-ray 83%
			Cost saving from omission of CXR	53 Canadian dollars/patient
Farah et al 2002 USA	140 healthy infants <12 months presenting with first episode of wheezing	Observational study	Rate of alternate diagnosis	0.7% (cardiac abnormality) 16%
			Rate of infiltrate/atelectasis	Fever -35%
			Severity and symptoms in children with atelectasis	Tachypnoea- 52% Hypoxaemia -39%
Shaw et al, 1991 USA	213 infants <13 months presenting to ED with bronchiolitis. Assessment of clinical and radiological features on initial presentation as predictors of more severe disease	Prospective observational study	Atelectasis on CXR as a predictor of severe disease	OR 2.7 (95% CI 0.97 to 3.70) Sensitivity 21% Specificity 97% PPV 82% NPV 70%	Disease severity dichotomised into ‘mild’ and ‘severe’. Radiologists blinded to clinical findings but investigators not blinded to CXR result

Comment(s)

Several of the studies include either older children or patients with previous episodes of wheezing and therefore may not be applicable to infants with bronchiolitis. Rate of alternative diagnosis in patients with typical bronchiolitis is low and in the majority of patients presenting with bronchiolitis x-ray is unlikely to influence management. Radiological changes do not correlate strongly with clinical severity. The studies by Friis et al (1990) and Eriksson et al (1986) also suggest that x-ray changes do not reliably differentiate patients with bacterial infection from those with viral infection. Overuse of chest radiography may increase the rate of false-positive diagnosis of ‘pneumonia’ and consequently unnecessary antibiotic use.

Clinical Bottom Line

Routine chest radiography is not indicated in infants presenting with simple bronchiolitis

Level of Evidence

Level 2 - Studies considered were neither 1 or 3.

References

1. Friis B, Eiken M, Hornsleth A, et al. Chest Xray appearances in Pneumonia and Bronchiolitis. Acta Paediatr Scand 79 1990 ; 219-225.
2. Schuh S, Lalani A, Allen U, et al Evaluation of the Utility of Radiography in Acute Bronchiolitis Journal of Pediatrics April 2007; 429-433
3. Mahabee-Gittens EM, Bachman D, Shapiro E, et al. Chest radiographs in the pediatric emergency department for children (less than or equal to) 18 months of age with wheezing Clinical Pediatrics July 1999; 395-9
4. Dawson K, Long A, Kennedy J, et al. The Chest Radiograph in Acute Bronchiolitis J. Pediatr Child Health 1990; 209-211
5. Eriksson J, Nordshus T, Carlsen KH, et al. Radiological findings in children with respiratory syncyctial virus infection: relationship to clinical and bacteriological findings. Pediatr Radiol 1986;16:120–2.
6. Yong J, Schuh S, Rashidi R, et al. A cost effectiveness analysis of omitting radiography in diagnosis of acute bronchiolitis. Pediatr Pulmonology 2009;44:122–7.
7. Farah M, Padgett L, McLario D, et al. First time wheezing in infants during respiratory syncytial virus season: chest radiograph findings. Pediatr Emerg Care 2002;18:333–6.
8. Shaw K, Bell L, Sherman N. Outpatient assessment of infants with bronchiolitis. Am J Dis Child 1991;145:151–5.