Applicability of the National Healthcare Safety Network's surveillance definition of ventilator-associated events in the surgical intensive care unit: a 1-year review.

BACKGROUND: In 2012, the National Healthcare Safety Network presented a new surveillance definition for ventilator-associated events (VAEs) to objectively define worsening pulmonary status in ventilated patients. VAE subcategories, ventilator-associated condition (VAC), infection-related VAC, and probable ventilator-associated pneumonia (PrVAP), were vetted predominantly in medical intensive care units. Our goal was to evaluate how well VAE criteria characterize pulmonary complications in surgical intensive care unit (SICU) patients.

METHODS: Since September 2012, all intubated SICU patients were screened prospectively for VAE and monitored for sustained respiratory dysfunction that did not meet VAE criteria. We diagnosed ventilator-associated pneumonia (VAP) using a clinical definition: Clinical Pulmonary Infection Score (CPIS) greater than 6 and catheter-directed bronchoalveolar lavage cultures with 10 or more colony-forming units per milliliter of pathogenic organisms.

RESULTS: We admitted 704 intubated patients. A total of 437 were intubated for two or more days (mean [SD], age 46 [18] years; 65% male; median ventilator days, 4 [range, 2-9]; median Sequential Organ Failure Assessment [SOFA] score, 8 [range, 5-10]). Using VAE criteria, we identified 37 patients with VAC, 31 with infection-related VAC, and 22 with PrVAP. While the remaining 400 patients did not meet VAE criteria, we identified 111 patients (28%) with respiratory deterioration and diagnosed 99 additional pneumonias. Of the 111 patients, 85 (77%) never had a period of stable/decreasing oxygenation, requiring elevated vent settings upon initiation of ventilation preventing them from meeting VAE criteria. Of the 99 pneumonia patients, 10% had sustained respiratory deterioration treated with elevations in mean airway pressure; they did not meet VAE criteria as the positive end-expiratory pressure or FIO2 was not elevated. Twenty-seven percent never had a period of stable/decreasing oxygenation. Fifty-eight percent had less than 2 days of respiratory deterioration. Agreement between PrVAP and clinical VAP was 77.3% (κ = 0.243, p < 0.001).

CONCLUSION: The applicability of the new National Healthcare Safety Network categories of VAE to critically ill surgery patients is limited. Agreement between PrVAP and clinical VAP in SICU patients is poor. Most surgical patients are not well categorized by this new definition; a better method of surveillance should be created for this patient population.

LEVEL OF EVIDENCE: Diagnostic study, level III.