Medical Center

Skip to content. | Skip to navigation

Personal tools

Home > Clinical Nutrition Services > Inpatient Clinical Nutrition Services > Digestive Health > E-journal Club > jun04.html


Nutrition Support Traineeship

e-Journal Club

June 2004


Summer is in full swing here at the foothills of the Blue-Ridge Mountains , and for us that means new physicians have just arrived, and our dietetic interns will be graduating soon.  We have had the honor of providing clinical nutrition training to two dietitians, Kazuko Hoshino & Nana Matsumoto, from Japan the past 3 weeks at UVAHS.   Dr. Hosoya from the Nutrition Care Management Committee in Japan sent them to learn more about clinical care and management practices.  It has been great to “compare notes” and acquire their perspective on aspects of patient care and nutrition support.

We have recently made all of our e-journal clubs available on our website to make it easier to access and refer to them.  http://www.healthsystem.virginia.edudh/traineeship.html

Now that our words reach a wider audience we wanted to let folks know we don’t intend this to be an exhaustive review (we also had to curb Joe’s freestyle form a bit….J L.  The goal of our e-journal club is to provide a brief summary, mention key points of our discussion, and encourage everyone to critically evaluate the article for yourself. 

Our article this month is from the May-June 2004 JPEN:

Shang E, Geiger N, Sturm JW, Post S.  Pump-assisted enteral nutrition can prevent aspiration in bedridden percutaneous endoscopic gastrostomy patients.  J Parenter Enteral Nutr. 2004 May-Jun;28(3):180-3.
University Hospital Mannheim , Department of Surgery, Mannheim , Germany .


This study compared two methods of feeding bedridden patients via a PEG.  100 patients were randomized to receive either pump-delivered feeds or gravity-drip feeding for 6 weeks, and then the groups were crossed-over to receive the other method of feeding.  All patients received a 1 calorie/ml formula.  Pump-delivered feedings were provided over a maximum of 12 hours (median 7.4 +/- 1.1 hrs) at a mean flow rate of 202 mL/hr, while the gravity-drip was delivered over 5.1 +/- 3.5 hrs (median time), mean flow rate of 306  mL/hr.  Average calorie delivery was 1140 calories/day – similar average estimated calorie needs between groups.

The primary outcomes measured were diarrhea, regurgitation, vomiting and aspiration, pneumonia, and blood glucose on days 1, 21, and 42 of the study.

The authors reported a significant difference in regurgitation, vomiting, aspiration, pneumonia and diarrhea with pump feedings, when compared to gravity feedings.  They also reported an “improved glucose profile” with pump feeding.


     This study has several strengths, including the crossover design, and the relatively large number of similar patients (N = 100 [50 per group]; 62-80 years, all bedridden patients at home on PEG feedings x 6 months prior). 

One primary limitation is that the study is not double-blind, and several of the primary outcomes could be influenced by subjective bias.  Diarrhea was defined as an increased frequency of BM to 2X “usual” or change to “watery form”.  The evaluation for pneumonia included the presence of cough, wheezing, or rales (in addition to an infiltrate).

     The other limitation is that there is no mention of how they determined or defined several outcomes such as regurgitation or aspiration.  They define their criteria for “aspiration pneumonia”, but report “aspiration” and “pneumonia” as separate entities in the table.  There is no mention of how often patients received a chest X-ray to detect the presence of an infiltrate, or if there were no standing or daily chest X-ray, there is no mention of the clinical criteria for a chest X-ray to be ordered.  There is no mention if these patients had the head of their bed consistently elevated during feeding, or how often they were monitored for proper positioning during feeding.

     One of the issues that we discussed is that it is amazing that no patient EVER had diarrhea, vomiting, aspiration, or pneumonia while they were receiving pump-feedings.  Given that patients were switched right over to the opposite method of feeding, it is just amazing how all problems abruptly ceased with the start of pump feeds, with not a single instance of carry over of even a day of diarrhea into the next phase!  In the area of glucose control, it is important to note that no patient in either group experienced hyperglycemia,…ever.  This appears to be one of those issues that is statistically significant, but was not clinically significant in this population.

     The primary focus of our groups concern was in the area of how applicable are these results to our patients (sick, hospitalized).  The gravity-fed patients received their complete nutrition needs in a single feeding, delivered over approximately 5 hours (median time).  In our practice, most patients who do not receive pump feedings have their bolus or gravity feedings distributed over 3-4 feedings per day.  In order for this study to have significant implications for our patients it would be necessary for patients to be randomized to bolus or gravity feedings spread out over 3-4 feedings per day, compared to pump-delivered feedings.

Of note, insurance companies are going to be more restrictive on the use of pumps due to the lack of evidence in the benefits over gravity drip.   

Take home message: 

This study’s results may not apply to patients who receive their nutrition in multiple feedings over the day.  The unblinded nature of the study presents opportunity for bias, but these results do suggest that providing complete nutrition in a very short period of time (2.6 to 8.6 hrs) may increase the risk of aspiration pneumonia in a bedridden, neurologically impaired population.

Finally, it is good to know that patients can be fed by gravity drip at home if need be; perhaps a flow rate of 300 mL/hr is not a good way to feed however. 

"Not everything that can be counted counts, and not everything that counts can be counted."
-Albert Einstein