Nine of the 13 studies reported data on infectious complications
with EN versus PN. The nature of the infectious complications
varied with the particular patient population and included pneu-
monia, aspiration pneumonia, urinary tract infections, bacteremia,
wound infection, abdominal abscess, and line sepsis. When the
data were aggregated from these studies (Figure 1), there was a
significant decrease in the number of patients with infectious
complications who had received EN rather than PN (RR ⫽ 0.64,
95% CI ⫽ 0.47 to 0.87). The test for heterogeneity of this aggre-
gate was not statistically significant (P ⫽ 0.22). All 13 studies
reported mortality rate as an outcome. The result of this analysis
(Figure 2) demonstrated no difference in mortality rate in critically
ill patients on EN versus PN (RR ⫽ 1.08, 95% CI ⫽ 0.70 to 1.65),
with a non-significant test for heterogeneity of 0.2.
When a subgroup of studies in which the PN group was fed
more calories than the EN group (non-isocaloric dosing across