Eighteen studies with a total of 2,677 patients were included: 15 RCTs and 3 quasi-randomised controlled trials.
The data were not pooled qualitatively as studies were considered to be heterogeneous.
Ventilator circuit changes (3 studies).
No significant differences were observed in any of the 3 studies for ventilator circuit changes compared with no change, or for changes every 7 days compared with 3 times per week. The RRs ranged from 0.84 to 0.91.
Heat humidifiers versus heat and moisture exchanges (5 studies).
The RRs ranged from 0.34 to 0.86. Only one study had a significant result (RR 0.41, 95% CI: 0.20, 0.86) in favour of heat humidifiers. From the results of this study, the authors calculated that 10 patients would have to be treated using heat and moisture exchangers instead of heat humidifiers, in order to prevent one case of VAP.
Oral versus nasal endotrachial entubation (1 study): the RR was 0.52 (95% CI: 0.24, 1.13).
Subglottic secretion drainage (2 studies).
One study had a non significant result (RR 0.57, 95% CI: 0.32, 1.01), whilst the other was significant at the p less than 0.05 level (RR 0.46, 95% CI: 0.23, 0.93). From the results of this study, the authors calculated that 7 patients would have to be managed using subglottic secretion drainage instead of a standard endotrachial tube, in order to prevent one case of VAP.
Closed versus open system suctioning (2 studies): neither study had a significant result; the RRs ranged from 0.90 to 0.95.
Kinetic bed therapy (3 studies).
The RRs ranged from 0.35 to 0.78. Only one study had a significant result (RR 0.35, 95% CI: 0.16, 0.75). From the results of this study, the authors calculated that 4 patients would have to be managed on a kinetic rather than a standard bed, in order to prevent one case of VAP.