Randomized trials in critical care nutrition: look how far we've come! (and where do we go from here?).

BACKGROUND

The purpose of this methodological review is to quantify and qualify critical care nutrition randomized controlled trials (RCTs) that inform our practice, to evaluate their strengths and limitations, and to recommend strategies for improving the design of future trials in this area.

METHODS

The literature was systematically reviewed to find all RCTs published between 1980 and December 2008 that evaluated nutrition interventions in critical care. Data were abstracted on the nature and quality of included RCTs.

RESULTS

A total of 207 RCTs met the inclusion criteria. Of these, 170 (82.1%) were single-center, and 37 (17.9%) were multicenter. The largest number of trials evaluated intensive insulin therapy (n = 25), arginine-supplemented diets (n = 22), and supplemental parenteral glutamine (n = 17). The first RCTs were published in 1983 (n = 2), and the mean sample size was 39.0. In 2008, there were 26 RCTs, each enrolling an average of 237.1 patients. Excluding 2 cluster RCTs, 62 of 205 (30.2%) trials had concealed randomization, 125 of 205 (61.0%) reported on intention-to-treat analyses, and 69 of 205 (33.7%) had a double-blinded intervention; 18 of 205 (8.8%) studies reported on all 3 design characteristics. Currently, 60 critical care nutrition RCTs (18 multicenter trials) are registered on clinical trials registries.

CONCLUSIONS

The future of clinical critical care nutrition research is promising, with more trials of increasing sample size being conducted. Robust trial methodology, transparent reporting, and the development of research networks will help to further advance this important field.