Background and Aims Ventilator-associated pneumonia (VAP) is associated with increased length of stay and adverse outcomes in PICU patients. In a retrospective study, we examined if PRISM III score at admission or at the day of VAP development could better predict the outcome in patients with VAP.

Methods The medical records of PICU patients admitted to a 8-Bed PICU of a tertiary-care hospital from January–December 2011 were reviewed. Clinical data, PRISM III score at admission or at the day of VAP development were recorded. VAP was diagnosed according to CDC criteria.

Results 27 patients, mean age 4.40±4.23 years, 59.3% boys, developed VAP. 4 patients presented 2 VAP episodes. Mean PRISM III score at admission was 10.19±7.65, at the day of first VAP episode 7.31±6.94, and at the day of 2nd VAP episode 4.75±3.60. The receiver operator characteristic curve (ROC) analysis showed that PRISM III at admission could better predict mortality in PICU patients with VAP than PRISM III at the day of VAP episode. (Figure 1).

• Download figure
• Open in new tab
• Download powerpoint

Abstract 811 Figure 1

The area under the curve was found 0.85 (asymptotic 95%CI 0.59 to 1, P<0.01) for PRISM III at admission and 0.72 (95%CI 0.48 to 0.95, P=0,081) for PRISM III at VAP episode.

Conclusions PRISM III at admission could better predict mortality in PICU patients with VAP than PRISM III at the day of VAP episode suggesting that VAP may not independently affect mortality.