Infections and Feeding Tubes

Sandy Beauman, MSN, RNC-NIC


While risk of infection is a constant concern in the neonatal unit, until recently the specific connection between feeding tubes and infection had not been made.  There is some evidence that the feeding tube and/or the feeding solution may be the culprit, particularly in adults.

Careful handling of expressed breast milk and formulas is critical and well-researched and advocated by the American Dietetic Association and the Human Milk Banking Association of North America (HMBANA).1, 2  Over the past few years, more attention has gone to the feeding tube itself, particularly as related to intestinal colonization and bacterial translocation.   Furthermore, in the California Central Line Blood Stream Infection Prevention project that was in progress for 8 years, in following all blood stream infections, we asked sites to do an intensive evaluation of every blood stream infection in order to determine, whenever possible, to what it might be related.  A surprising percentage, 15%, were related to “gut issues” which may contribute to bacterial translocation, feeding intolerance, necrotizing enterocolitis and ultimately, sepsis.

There are two published studies that have looked specifically at colonization of feeding tubes, one of which also looked at blood stream infections related to the specific organism cultured from the feeding tube.  These studies evaluated bacteria found in the feeding tube based on the feeding solution and dwell time of the tube.3, 4   The Hurrell study’s intent was to identify any association with feeding regimen i.e. formula vs breast milk.  They found colonization occurred in all feeding tubes, even those from babies who were not fed through the tube.  The organisms were different but still present.  Two of the organisms caused infections in the NICUs.  These two organisms accounted for 51% of all organisms and were Klebsiella and Serratia.  The Mehall study cultured tubes after 7 days and found all but one tube was colonized with various organisms, including Klebsiella.  Based on these studies, some institutions have shortened the dwell time of the feeding tube to no more than 7 days.  However, the Mehall study did not document that there were no organisms prior to 7 days, just that they were present at 7 days.  In an adult study, the colonization of the feeding tube was traced to the gut vs the feeding solution.5  This group documented retrograde travel of the organisms into the feeding tube.  Mehall et al showed this as well but also found organisms that are not present in the gastrointestinal tract, suggesting contamination from another source.  This was posited to be from healthcare workers or the environment and not from the feeding solution itself.  In the study by Hurrell and his group, tubes were cultured that had been in place for less than 6 hours up to 48 hours.  Bacterial growth was highest in the tubes that had been in place longer, with all tubes in place longer than 6 hours having significant growth.  While there may not be any definitive answer about how long feeding tubes should be in place, it seems obvious that a shorter dwell time will be better for the baby than a longer dwell time.  Practice and manufacturer’s recommendations vary from changes with every feeding to changing the tube every 7 days for short term feeding tubes and 30 days for long term feeding tubes.  Neither of the studies discussed here mentioned the material the feeding tube was made from so at this point, there seems no clear link between the material and bacterial colonization or biofilm formation.

So, if the feeding tube itself should be changed due to biofilm formation and presence of bacteria, how often should extension tubing be changed?  This is usually changed more often at a 4-hour interval in many cases.  This is based on ADA and HMBANA recommendations.  In some cases, longer feeding tubes are used in order to prevent use of extension tubing.  This may be in an effort to avoid one more supply needed, to avoid this frequent tubing change or for other reasons.  We know from performance improvement efforts related to central line associated blood stream infections that more connections increase the risk of infection.  So, one may pose that a single tube (the feeding tube) that is long enough to serve the purpose of the extension tube and feeding tube would be advantageous.  However, the use of this longer feeding tube may lead to additional reservoirs in the tubing where bacteria may adhere and more difficulty in clearing the tube, even when flushed.

Speaking of flushing, the practice in adults is to flush the tube to remove formula and not allow it to linger in the tube at body temperature which would provide a nice warm environment and nutrition for bacteria to grow.  The purpose of this flush is as much to prevent obstruction of the feeding tube as to remove the environment for bacterial growth.  This practice is being adopted more in neonates but when the tubes are flushed, they may be flushed with air or water.  The amount of the flush is very small even when proportions are considered with the adult population as there is concern about administering too much water to the infant.  This begs the question about how well the tube is cleared, even when “flushed”?  In addition, neonates are often fed continuously, not allowing for flushing of the tube and constant movement of formula or expressed breast milk through the tube.  This constant movement may decrease the bacterial growth but this has not been studied specifically.  Neither has bacterial growth or biofilm formation been studied in gastric tubes that are routinely flushed.

Berthelot et al reported on an infectious outbreak in France.6  This included digestive tract colonization of Klebsiella  oxytoca in 24 infants.  One of these infants died and others were clinically ill.  Ultimately, the colonization and cross-contamination was eradicated with the implementation and enforcement of hand hygiene and use of gloves when handling feeding tubes.  Gastmeier et al reported a summary of regarding outbreaks in NICUs.7  They identified Klebsiella as the most frequent organism involved in reported neonatal outbreaks (Keep in mind that this represents a fraction of all outbreaks, in all likelihood.)  Most of the neonatal outbreaks were not able to be traced to any specific source in most cases while adult outbreaks often were but Klebsiella was found in both studies discussed earlier when feeding tubes were cultured.  Therefore, the attention to detail and perhaps moving toward a safer practice without additional evidence is warranted in this case.  At least, there seems to be sufficient evidence to support changing feeding tubes at some defined interval not to exceed 7 days, using clean gloves for handling feeding tubes and maintaining the feeding tube as a closed system as much as possible.


1. The Human Milk Banking Association of North America, Inc. (2011).   Best Practice for Expressing, Storing and Handling Human Milk in Hospitals, Homes and Child Care Settings. HMBANA: West Hartford, CT.

2. Robbins S, Meyers R. Pediatric Nutrition Practice Group. (2011).  Infant feedings: Guidelines for preparation of human milk and formula in health care facilities 2nd ed.  American Dietetic Association.

3. Hurrell E, Kucerova E, Loughlin M, Caubilla-Barron J, Hilton A, Armstrong R, et al.  Neonatal enteral feeding tubes as loci for colonization by members of the Enterobacteriaceae.  BMC Infectious Diseases.  2009; 9:146.

4. Mehall JR, Kite CA, Saltzman DA, Wallett T, Jackson RJ, Smith SD.  Prospective study of the incidence and complications of bacterial contamination of enteral feeding in neonates.  Journal of Pediatric Surgery.  2002; 37(8):1177-1182.

5. Mathus-Vliegen EMH, Brediu MWJ, Binnekade JM.  Analysis of sites of bacterial contamination in an enteral feeding system.  Journal of Prenteral and Enteral Nutrition.  2006;30(6):519-525.

6. Berthelot P, Grattard F, Patural H, Ros A, Jelassi-Saoudin H, Pozzetto B, et al. Nosocomial colonization of premature babies with Kelbsiella Oxytoca: Probable role of enteral feeding procedure in transmission and control of the outbreak with the use of gloves.  Infection Control and Hospital Epidemiology.  2001;22:148-151.

7. Gastmeier P, Loui A, Stamm-Balderjahn S, Hansen S, Zuschneid I, Sohr D, et al.  Outbreaks in neonatal intensive care units—they are not like others.  American Journal of Infection Control.  2007; 35:172-176.


Read more of Sandy’s infection-related NICU blog posts.
View her blog entry Fat Loss During Tube Feedings.

Are you looking for a feeding tube solution for your NICU?
Learn more about the Medela Enteral Feeding System feeding tubes.


About the Author

Sandy Sundquist Beauman has over 30 years of experience in neonatal nursing. In addition to her clinical work, she is very active in the National Association of Neonatal Nurses, has authored or edited several journal articles and book chapters, and speaks nationally on a variety of neonatal topics. She currently works in a research capacity to improve healthcare for neonates. Sandy is also a clinical consultant with Medela. You can find more information about Sandy and her work and interests at