Colom and Teper defined a score for diagnosing post-infectious bronchiolitis obliterans in children. The authors are from R. Gutierrez Children's Hospital of Buenos Aires in Argentina.
Patient selection: pediatric patient under 2 years of age
Outcome: post-infectious bronchiolitis obliterans
Parameters:
(1) adenovirus infection
(2) typical clinical history (previously healthy infant; severe episode of bronchiolitis; chronic respiratory hypoxemia for > 60 days; oxygen saturation < 92% at rest and breathing room air)
(3) high-resolution chest CT (HRCT) showing a mosaic pattern
(4) need for mechanical ventilation
|
Parameter
|
Finding
|
Coefficient
|
Points
|
|
adenovirus
|
absent
|
0
|
0
|
|
|
present
|
2.9
|
3
|
|
clinical history
|
absent
|
0
|
0
|
|
|
present
|
3.8
|
4
|
|
HRCT mosaic
|
absent
|
0
|
0
|
|
|
present
|
3.7
|
4
|
|
MV requirement
|
absent
|
0
|
NA
|
|
|
present
|
1.3
|
NA
|
X =
= SUM(coefficieints for all 4 parameters) - 2.9
probability of bronchiolitis obliterans =
=1 / (1 + EXP((-1) * X))
prediction rule =
= SUM(points for the first 3 parameters)
Interpretation of clinical rule:
• minimum score: 0
• maximum score: 11
• Possible scores: 0, 3, 4, 7, 8, 11.
• A score >= 7 was associated with bronchiolitis obliterans.
• A score of 7 would be achieved if any 2 of the risk factors are present.
Performance:
• The area under the ROC curve for the prediction rule in the derivation set was 0.96 (which seems high for a prediction rule). The specificity was 97% but sensitivity 76%.
