Three Part Question
In [a 5 year old girl with febrile neutropenia] does [routine chest radiography] assist in [management decisions or diagnose occult pneumonia]?
Clinical Scenario
A friendly, coryzal 5 year old girl with acute lymphocytic leukemia attends with another episode of febrile neutropenia. According to departmental protocol, her admission includes a chest x-ray. You wonder as to the value of this routine irradiation.
Search Strategy
SumSearch
"neutropenia" AND "radiography" AND filter "diagnosis"
Search Outcome
67 individual articles, 3 relevant.
Relevant Paper(s)
Author, date and country |
Patient group |
Study type (level of evidence) |
Outcomes |
Key results |
Study Weaknesses |
Feusner J et al, 1988, USA | 64 paediatric oncology patients with 134 episodes of febrile neutropenia | Prospective cohort | Prevalence of 'infectious' infiltrates | 3.0% (95% CI 0.81-7.7%) | 'Infectious' infiltrate was diagnosed partly by clinical findings - may exaggerate usefulness of clinical signs |
Diagnostic usefulness of 'tachypnoea, chest pain or abnormal ausculation' | LR+82 (95% CI 11-575), LR-0.0 (95% CI 0.0-0.19) |
Katz JA et al, 1991, USA | 131 paediatric oncology patients with febrile neutropenia | Prospective cohort | Prevalence of pneumonia | 3.1% (95% CI 0.7-7.8%) | Only 128/131 patients received radiographs |
Diagnostic usefulness of presence of respiratory signs | LR+ infinite (lower 95% CI 0.00), LR-0.5 (95% CI 0.25-1.0) |
Korones DN et al, 1997, USA | 54 paediatric oncology patients with 108 episodes of febrile neutropenia | Prospective cohort | Prevalence of pneumonia | 3.7% (95% CI 0.14-7.2%) | Only 40/54 patients received a chest radiograph |
Diagnostic usefulness of 'abnormal auscultatory findings, RR>20 when afebrile or O2 sats <95% twice in 4 hours' | LR+ 17.3 (95% CI 7.9-38), LR-0 (95% CI 0-0.79) |
Comment(s)
There is no good quality study addressing the use of chest radiographs in uncomplicated febrile neutropenia. Two of these studies are consistent with clinical feeling - lack of abnormal signs or symptoms in children with febrile neutropenia rules out pneumonia. The methodological weaknesses would tend to favour this - with one study having clinical features as part of the reference standard, and the second tending to fail to perform chest radiography on children without symptoms. The third study only gives data on respiratory signs (ignoring symptoms) and has a subsequently reduced sensitivity and improved specificity.
Clinical Bottom Line
Pneumonia is uncommon in children with febrile neutropenia (3%). An absence of respiratory signs and symptoms makes pneumonia very unlikely. Routine chest x-rays seem unnecessary.
References
- Feusner J, Cohen R, O'Leary M et al. Use of routine chest radiography in the evaluation of fever in neutropenic pediatric oncology patients. J of Clin Oncol 1988;6(11):1699-702.
- Katz JA, Bash R, Rollins N et al. The yield of routine chest radiography in children with cancer hospitalized for fever and neutropenia. Cancer 1991;68(5):940-3.
- Korones DN, Hussong MR, Gullace MA. Routine chest radiography of children with cancer hospitalized for fever and neutropenia: is it really necessary? Cancer 1997;80(6):1160-4.