Introduction
The best method for diagnosis of ventilator-associated pneumonia (VAP) in children is much debated. Clinical criteria alone may not be sufficiently reliable. Bronchoscopic techniques are not routinely used. Blind (nonbronchoscopic) protected bronchoalveolar lavage (BAL) has been studied in pediatrics but has never been validated according to the gold standard (autopsy).
Objective
To compare different diagnostic methods of VAP in children, using clinical judgment of an adjudicating committee of experts as the reference standard.
Methods
Setting
Prospective study of all consecutive PICU patients < 18 years with suspected VAP.
Diagnostic methods compared
1) clinical data using Centers for Disease Control (CDC) criteria; 2) blind protected BAL, evaluating quantitative cultures, bacterial index, Gram stain and presence of intracellular bacteria; 3) nonquantitative cultures of endotracheal secretions.
Reference standard
Consensus of three independent experts (Delphi method) who retrospectively established the presence of VAP based on clinical, microbiological and radiological data.
Analysis
Concordance between each diagnostic method and the reference standard was evaluated by concordance percentage and kappa score. Validity was evaluated using sensitivity, specificity and global value.
Results
Thirty patients were included in the study. According to the reference standard, VAP occured in 10/30 patients (33%). Concordance and validity of the different methods are presented in Table 1.
Table 1.
| Diagnostic method | Number of patients with positive results (n/30) | Concordance with reference standard (%) | Kappa score | Sensitivity (%) | Specificity (%) | Global value (%) |
|---|---|---|---|---|---|---|
| Clinical criteria (CDC) | 28 | 43 | 0.11 | 100 | 15 | 43 |
| BAL | ||||||
| Culture ≥ 102 CFU/ml | 14 | 67 | 0.32 | 70 | 65 | 67 |
| Culture ≥ 103 CFU/ml | 11 | 70 | 0.34 | 60 | 75 | 70 |
| Culture ≥ 104 CFU/ml | 9 | 70 | 0.31 | 50 | 80 | 70 |
| Culture ≥ 105 CFU/ml | 6 | 67 | 0.17 | 30 | 85 | 67 |
| Intracellular bacteria | 4 | 73 | 0.29 | 30 | 95 | 73 |
| Gram stain | 9 | 70 | 0.31 | 50 | 81 | 73 |
| Bacterial index > 5 | 9 | 83 | 0.61 | 78 | 86 | 90 |
| Endotracheal culture | 21 | 57 | 0.24 | 90 | 40 | 57 |
Conclusion
Our data show that the most reliable diagnostic method for VAP is a bacterial index > 5. Further studies should evaluate the validity of all these methods according to the gold standard (autopsy).