Make the Diagnosis: Pneumonia, Ventilator-Associated
Ventilated patients are highly susceptible to pneumonia with pooled cumulative incidence of 9.7% of ventilated patients in a meta-analysis of 38 units published in 2005.1 This incidence was 9.1% in combined medical-surgical units and 17% in medical.1 In 2008, the US Joint Commission launched a multiyear effort to decrease the incidence of ventilator-associated pneumonia.2 This may translate to lower rates in individual units and clinicians are encouraged to be aware of their local improvement efforts to have an understanding of the incidence in their intensive care unit. In addition, the probability of ventilator-assisted pneumonia likely increases with days of ventilator support, so individual patient risk should be modified for length of intubation.
Patients for Whom Ventilator-associated Pneumonia Should Be Considered
All mechanically ventilated patients are at risk for pneumonia. The clinical challenge is that the signs of pneumonia are present in other illnesses that commonly affect this population including acute respiratory distress syndrome, thromboembolic disease, alveolar hemorrhage, sepsis, congestive heart failure, and atelectasis. However, the gradual development of fever, tachypnea, and hypoxemia are believed to be harbingers of ventilated-associated pneumonia.
Assessing the Likelihood That a Ventilated Patient Has Pneumonia
Ventilated patients should be assessed for a combination of findings. While individual clinical findings have limited value for identifying the patient with pneumonia (see Table 62-1), the absence of a new radiographic infiltrate makes pneumonia less likely. Combinations of findings are more useful for identifying patients with pneumonia, but the confidence intervals are broad. The analysis of pulmonary secretions from bronchoalveolar lavage has the most useful likelihood ratios, though some of the associated confidence intervals are broad.
Table 62-1Likelihood Ratios That a Ventilated Patient Has Developed Pneumonia |Favorite Table|Download (.pdf) Table 62-1 Likelihood Ratios That a Ventilated Patient Has Developed Pneumonia
| ||Likelihood Ratios |
|Findings ||LR+ (95% CI) ||LR- (95% CI) |
|Clinical Findings |
|New radiographic infiltrate ||1.7 (1.1-2.5) ||0.35 (0.14-0.87) |
|Abnormal white blood cell count ||1.3 (0.76-2.4) ||0.74 (0.34-1.6) |
|Sputum production ||1.3 (0.88-1.8) ||0.63 (0.28-1.4) |
|Fever ||1.2 (0.76-1.9) ||0.86 (0.54-4.1) |
|Pulmonary Secretions |
|High cellularity (>400 000 cells/mL) in BAL fluid3 ||15 (2.3-103) ||0.11 (0.03-0.40) |
|>50% neutrophils in BAL fluid4 ||2.0 (1.4-3.0) ||0.09 (0.01-1.4) |
|Combinations of Findings |
|>2 of fever, leukocytosis, and purulent sputum5 ||2.8 (0.97-7.9) ||0.41 (0.17-0.99) |
There is a surprising lack of consensus on the criteria for ventilator-associated pneumonia. In this review, studies were included only if there was a histological examination of lung tissue from a lung biopsy or a postmortem examination. However, this approach creates more consistency than reference standards that include bronchoscopic culture that depend on technique of the bronchoscopist and the criteria for culture positivity.