Ten studies with a total of 583 patients were included. There were 4 studies (total n=265) of EN compared with TPN, one trial (n=38) of early versus delayed nasoduodenal feeding, one trial (n=38) of gastric versus jejunal tube feeding, one trial (n=60) of intermittent versus continuous enteral feeding, and 3 trials (total n=181) evaluating different enteral feeding formulae.
The four trials comparing TPN with EN yielded inconsistent results. In one trial, there was a trend towards a lower rate of pneumonia associated with EN (0% compared with 20% in the TPN group; RR undefined), while in another, the pneumonia rate was significantly lower in the EN group (RR 0.38, 95% CI: 0.16, 0.90). In the remaining two studies, the pneumonia rate was slightly higher in patients receiving EN (RR 1.23, 95% CI: 0.51, 2.95; RR 1.06, 95% CI: 0.56, 2.02), but this was not statistically significant.
One study compared early nasoduodenal feeding begun within 24 hours with nasoduodenal feeding delayed for 72 hours. In patients receiving early feeds, there was a non significant trend towards increased rates of pneumonia; the rates of pneumonia were 42 and 21% for early and late feeds, respectively (RR 2.0, 95% CI: 0.72, 5.54).
Considering the potential for EN to cause aspiration pneumonia, one study tested the effect of the delivery site. Two cases of pneumonia were identified among the 19 patients receiving pre-pyloric gastric feeds, while no cases were observed in the 19 patients receiving post-pyloric feeds through a jejunal tube (RR undefined).
In the single trial comparing intermittent and continuous EN, 5 of the 30 patients in each group developed nosocomial pneumonia (RR 1, 95% CI: 0.32, 3.10).
One study evaluated a MTF, i.e. a high-protein, low-fat, linoleic acid-restricted formulation enhanced with arginine, cysteine, vitamin A, zinc, omega-3-polyunsaturated fatty acids, and ascorbic acid. When compared against standard preparations (Traumacal and Osmolite), lower pneumonia rates were observed in the MTF patients. This difference was statistically significant when comparing Traumacal with MTF (RR 0.25, 95% CI: 0.06, 0.99), but statistically non significant when comparing Osmolite with MTF (RR 0.27, 95% CI: 0.07, 1.15).
There was no difference in the rates of pneumonia when Immun-Aid was compared with Vivonex (RR 0.92, 95% CI: 0.24, 3.48). However, when Immun-Aid was compared with Promote, it was found to be associated with a non significant trend towards a lower pneumonia rate (0% compared with 18% in the Promote group; RR undefined).