University of Virginia Nutrition
Summer has arrived ! At UVAHS, this means our ‘senior' GI fellows have departed and we will be welcoming new GI fellows and new resident physicians to the hospital. June is a month of many transitions, school is out, and (most importantly) camping, canoeing and vacations are in - so we are unable to host a traineeship in June. However, the progress of Science waits for no one - so the Nutrition Support Journal Club must go on!
- Artinian V, Krayem H, DeGiovine B. Effects of Early Enteral Feeding on the Outcome of Critically Ill Mechanically Ventilated Medical Patients. Chest 2006;129:960-967.
This is an observational study of 4049 nonsurgical patients that required mechanical ventilation for at least two days. The data was extracted from a multi-institution database (Project Impact Critical Care Data System) in 2003. The investigators divided the patients into two groups based on when their enteral feeding was started. The early feeding group was started on EN within 48 hours of initiation of mechanical ventilation, and the late feeding group was defined as everyone else. Patients that died or were extubated within 2 days of starting mechanical ventilation were excluded. Patients were excluded if they had a diagnosis on admission, which would be a contraindication to enteral feeding, or if they had received parenteral nutrition prior to mechanical ventilation.
Variables collected included age, gender, race, diagnosis, mortality prediction model, Simplified acute physiology score (SAPS) II, acute physiologic and chronic health evaluation (APACHE) II, and ventilator associated pneumonia (VAP).
Logistic regression was used to assess the effect of early feeding on ICU mortality, hospital mortality and VAP. In addition, the early fed patients were matched with late fed patients in pairs via a "propensity score for the likelihood of being fed," developed using logistic regression. Analysis was also completed on this matched data set. The primary outcomes were ICU and hospital mortality, and the secondary outcome variables were VAP, ICU length of stay, and ventilator-free days.
Major Results reported by authors:
The authors reported that overall ICU and hospital mortality were significantly lower in the early feeding group. They also reported that the association between early feeding and improved survival was most pronounced in the sickest patients. The investigators stated that the significant association between early feeding and improved survival remained in the 1,264 pairs of patients that were matched via a "propensity score". The other result that was reported in every analysis that adjusted for severity of illness was that early enteral feeding was associated with increased risk of VAP.
The author's conclusions, as written in the abstract were "Early feeding significantly reduces ICU and hospital mortality based mainly on improvements in the sickest patients, despite being associated with an increased risk of VAP developing."
The group had strong objection to the way that the conclusions were written. This is a retrospective, observational study, and as such, one can only identify associations from the data. Regardless of the statistical control that is attempted, it is absolutely inappropriate to make cause and effect conclusions from an observational study. Outside of randomizing large numbers of patients, there is no way to adequately control for the myriad of factors that influence patient outcome in an ICU population. Nutritional status, for instance, was not controlled for in any way; nor amount, type or route of enteral feeding. GI motility is not factored into any of the illness severity scores that were used, although poor gut function has been associated with poor outcome in critically ill patients (Dunham). Patients with signs of gut dysmotility would be more likely to have a delayed start of enteral feeding.
Another factor that was not discussed, nor controlled for, was the start of TPN after the patients were receiving mechanical ventilation. The protocol states that patients were excluded from the study if they had received TPN before mechanical ventilation, but it does not exclude the possibility of starting TPN once mechanical ventilation began. Carol contacted the researchers and found out, in fact, (although this was not stated in the paper) that 393 of the 1512 patients (21%) in the late-feeding group were started on TPN, while only 193 of the 2537 patients (7.6%) in the early-feeding group were started on TPN. This difference in TPN use (and thus potentially glucose control) is not accounted for in their methods.
Take home message:
It is inappropriate to make cause and effect conclusions from an observational study. There are a number of factors that are not controlled for in this study. The results of this study do support the need for a large, randomized study to investigate the potential benefits, and risks, of early enteral feeding. A randomized study to answer the question of the effects of early enteral feeding should be designed, without the confounding factor of TPN.
Dunham CM, Frankenfield D, Belzberg H, et al. Gut failure--predictor of or contributor to mortality in mechanically ventilated blunt trauma patients? J Trauma 1994;37(1):30-4.
Check out the latest Practical Gastroenterology article:
Kossoff EH, Turner Z. The Ketogenic and Atkins Diets: Recipes for Seizure Control. Practical Gastroenterology 2006; XXX(6):53.
Scroll down to GI Nutrition in the far left column and click on link
Then scroll down to box with links within the nutrition site
Nutrition Articles in Practical Gastroenterology is in the right column.
Joe Krenitsky MS, RD
Carol Parrish RD, MS
PS - Please feel free to forward this on to friends and colleagues.