Permissive Underfeeding in the ICU

Posted by • June 17th, 2015

Nutrition among critically ill patients is widely considered important, but the ideal caloric targets remain a subject of debate.  Some believe higher caloric intake is helpful and can reduce mortality; others argue the exact opposite, pointing to studies linking caloric restriction to lower morbidity, as long as protein intake is adequate. This debate has prompted investigation of an underfeeding strategy as a way to reduce mortality in critically ill patients.

The Permissive Underfeeding versus Target Enteral Feeding in Adult Critically Ill Patients (PermiT) trial enrolled close to 900 critically ill adults in seven centers in Saudi Arabia and Canada. These patients were randomized to either standard feeding (70-100% of calculated caloric requirements) or to underfeeding (40-60% of caloric requirements) for up to two weeks. The various centers delivered enteral feeding according to their own protocols; calculated caloric intake also accounted for calories from parenteral nutrition, intravenous dextrose, and propofol (1.1 kCal per milliliter).  Protein intake was kept the same for the two groups, with the underfeeding group receiving additional protein supplements as well as saline or water to match the protein amount and volume received by the standard feeding group.  The primary outcome was 90-day mortality.  The investigators predicted an 8% absolute risk reduction in favor of the underfeeding group.

As intended, patients in the underfeeding group consumed fewer calories than those in the standard feeding group (average intake was 46% vs 71% of daily requirements). But the study found no difference in 90 day mortality between the two groups (29% in the standard group, 27% in the underfeeding group; P=0.58).  There were also no differences between the groups for a number of secondary outcomes, including length of ICU stay, in-hospital mortality, 28-day mortality, and 180-day mortality. Further, based on limited subgroup analyses, the study did not identify any subpopulations with differences in mortality between the two strategies.

“The collective results of our study and the two previous trials add to a growing body of research that suggests that standard feeding goals in critically ill patients do not improve clinical outcomes,” the authors write.

While underfeeding did not demonstrate a mortality benefit in this study, the authors note that the study was powered to detect an eight percent risk reduction, which means smaller treatment effects cannot be ruled out.  They also observe that some of the enrolled patients, particularly in the standard feeding group, failed to reach their target caloric intake, which would have decreased the gap in caloric intake between the two groups. Finally, less than fifteen percent of ICU patients who were screened for the study were ultimately enrolled, suggesting the need for caution before generalizing these results to other critically ill patients.

How do you determine nutrition goals for critically ill patients?  Have you seen a role for permissive underfeeding in your management of certain patient populations? 

Comments are closed.