Does ventilator-associated event surveillance detect ventilator-associated pneumonia in intensive care units? A systematic review and meta-analysis

BACKGROUND Ventilator-associated event (VAE) is a new surveillance paradigm for monitoring complications in mechanically ventilated patients in intensive care units (ICUs). The National Healthcare Safety Network replaced traditional ventilator-associated pneumonia (VAP) surveillance with VAE surveillance in 2013. The objective of this study was to assess the consistency between VAE surveillance and traditional VAP surveillance. METHODS We systematically searched electronic reference databases for articles describing VAE and VAP in ICUs. Pooled VAE prevalence, pooled estimates (sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV)) of VAE for the detection of VAP, and pooled estimates (weighted mean difference (WMD) and odds ratio ([OR)) of risk factors for VAE compared to VAP were calculated. RESULTS From 2191 screened titles, 18 articles met our inclusion criteria, representing 61,489 patients receiving mechanical ventilation at ICUs in eight countries. The pooled prevalence rates of ventilator-associated conditions (VAC), infection-related VAC (IVAC), possible VAP, probable VAP, and traditional VAP were 13.8 %, 6.4 %, 1.1 %, 0.9 %, and 11.9 %, respectively. Pooled sensitivity and PPV of each VAE type for VAP detection did not exceed 50 %, while pooled specificity and NPV exceeded 80 %. Compared with VAP, pooled ORs of in-hospital death were 1.49 for VAC and 1.76 for IVAC; pooled WMDs of hospital length of stay were -4.27 days for VAC and -5.86 days for IVAC; and pooled WMDs of ventilation duration were -2.79 days for VAC and -2.89 days for IVAC. CONCLUSIONS VAE surveillance missed many cases of VAP, and the population characteristics identified by the two surveillance paradigms differed. VAE surveillance does not accurately detect cases of traditional VAP in ICUs.