Jejunal Tube Feeding –ESPGHAN Position Paper

A recent position paper (IJ Borekaert et al. JPGN 2019; 69: 239-58) makes 33 recommendations on the use of jejunal tube feedings.

Full Text Link: The Use of Jejunal Tube Feeding in Children: A Position Paper by the Gastroenterology and Nutrition Committees of the European Society for Paediatric Gastroenterology, Hepatology, and Nutrition 2019.

Here are a few of the recommendations:

  • #1 Jejunal feeding is route of choice for enteral nutrition with failure of oral and intragastric feeds or gastric outlet obstruction
  • #5 Expert group recommends the use of jejunal feeding in children with acute pancreatitis only in cases in which oral or gastric feeding is not tolerated
  • #6 Recommends trial of continuous gastric feeds or a hydrolyzed or elemental formula prior to jejunal feedings
  • #8 & 9 Expert group recommends to consider UGI/SBFT and an upper GI endoscopy in all patients before jejunal tube placement
  • #12 Recommends NOT to use jejunal tube feedings in preterm infants (<37 weeks gestation).  This is based on systemic reviews including Cochrane review which concluded “that there is no evidence of any benefit for transpyloric feeding in preterm infants compared to gastric feeding”
  • #21 Recommends monitoring for nutrient deficiencies –checking copper, zinc, selenium, and iron every 6-12 months (Low level of evidence).  The authors note that some studies have shown reductions in these nutrients; this may be related in part to be due to bypassing the duodenum
  • #24 Avoiding using jejunal tube for medication unless absolutely essential or delivery into the stomach is not possible

Hood River Bridge (crossing Columbia River). Hood River, OR


Predicting the Need for Gastrojejunostomy Tube Placement Instead of Gastrostomy Tube Placement

A recent study (ME McSweeney et al. JPGN 2018; 66: 887-92) determined that preoperative characteristics were unable to determine which patients who had gastrostomy tube (GT) placement would ultimately need conversion to gastrojejunostomy (GJ) placement.

This retrospective study matched 79 GJ patients with 79 GT patients.

Key points: 

  • These patients had similar rates of successful preoperative nasogastric feeding trials (GT 84.5% vs GJ 83.1%), and similar rates of abnormal swallow studies (53.8% and 62.2% respectively).
  • In the entire cohort, 11 patients had fundoplication (all GJ patients)
  • GT patients were more likely to have tube permanently removed: 20.5% vs 2.5% for GJ patients. Many (45.6%) of the GJ converted patients went back to GT feeds
  • Overall, from an initial cohort of 902 patients, 8.8% “required conversion” to GJ feeds
  • GJ-converted patients had a trend towards fewer hospitalizations.

While not a result in the study, the issue of GJ compared with fundoplication is briefly discussed.  The authors in their discussion of preoperative workup state that

“the complications of fundoplication are more significant and the risks are higher than GJ placement”

In my view, this is one of the most consequential parts of their discussion.  While the authors have extensive experience, I think the issue regarding GJ tube placement and fundoplication is more murky.  GJ tubes can be difficult to maintain and I have not seen long-term well-controlled studies comparing outcomes between GJ placement and fundoplication.

Other pointers in the discussion:

  • Fundoplication has “minimal impact/no impact” to reduce respiratory-related admissions, mainly because the main mechanism is aspiration rather than reflux
  • For isolated oropharyngeal dysphagia, one could argue that “an enteral tube is not indicated anymore” based on published data

My take: This is an important retrospective study that illustrates how difficult it is to know preoperatively which patients need GJ placement (or fundoplication) compared to GT placement alone.  In our institution, we are reluctant to place GT placement if a patient has not demonstrated tolerance of nasogastric feeds.

Related blog posts:

Amelia Island

Best gastrostomy tube

A recent report touts the feasibility of a one-step percutaneous gastrojejunostomy (GJ) as the latest advance in enteral access (JPGN 2012; 54: 820-21).  This reference describes a new variation in technical placement: gastropexy using t-fasteners to secure gastrostomy tube site and then advancing neonatal scope via gastrostomy site to advance guidewire for  GJ placement.  This technique was used in three infants.

Most centers have developed their own protocols for enteral access and it is likely that the familiar approach to that center will be safest for their patient population.  Recently, the subject of gastrostomy tube placement was extensively reviewed in our institution (see below) due to variation in care at two children’s hospitals.  In one hospital, the surgical group primarily placed laparascopic button gastrostomies and argued that better visualization led to lower complications like colonic interposition.  Furthermore, this approach was considered similar in cost effectiveness as the group would place a primary Mic-Key® ( thereby eliminating the need for anesthesia for a button placement.

The alternative approach utilized a Corflo® gastrostomy tube (  The advantages of this approach were 1) less anesthetic time/a smaller operation, and 2) lower likelihood of tube dislodgment.  This group approach argued that dislodgment was the greatest risk and that there was no urgency for a button tube.

Despite a joint meeting of these groups weighing the pros and cons, there was not a single best gastrostomy tube.

My experience is that tube dislodgment is quite common with button tubes.  In addition, primary button tubes can be difficult to size when the patient is under anesthesia.  As such, it is my practice to discourage primary gastrostomy button placement.  In addition, most patients who need gastrostomy tubes can wait until they are good surgical candidates both in terms of cardiorespiratory status and size.


Gtube Products:
• AMT clamp –helps eliminate tubing pullouts

• Gtube washable pads  (specific web address:

Additional references:

  • -J Pediatr 2011; 159: 602. Preemptive gtube assoc with improved survival post Norwood. High number needed fundoplication.
  • -JPGN 2011; 53: 293. 95% success with PEG in infants 2.1-5.6kg
  • -JPEN 2011; 35: 50-55. Predictive factors of mortality after PEG.
  • -JPGN 2009; 49: 237. Gtube improves height & weight in Rett syndrome.
  • -Clin Gastro & Hep 2007; 5: 1372. PEG placement does NOT prolong life in dementia patients.
  • -Arch Dis Child 2006; 91: 478-82. PEG reduced hospitalizations for respiratory dz in 57 severely impaired children
  • -J Pediatr 2006; 149: 837. inreased risk of PEG in SMA type 1 -42% w aspiration; 17% death (2/12)
  • -Pediatrics 2004; 114: 458-61. Moratlity rate of 0.4% -one death related to sepsis/peritonitis & 5% complication rate.
  • -Teitelbaum JE, Gorcey SA, Fox VL. Combined endoscopic cautery and clip closure of chronic gastrocutaneous fistulas. Gastrointest Endosc. 2005;62(3):432-435
  • -JPGN 2006; 43: 624. Satisfaction with PEGs: 94% of parents viewed PEG as positive influence on child’s situation & 98% would have chosen PEG insertion again (n=121).
  • -Sullivan PB, Dev Med Child Neurol 2005; 47: 77-83. 57 CP pts -almost all had improved health/nutrition p gtube
  • -Gastroenterology 2001; 121: 970-1001 & JPEN 2004; 28: S16. Provision of nutrition does not, for the most part, favorably alter clinical outcome.
  • -Lancet 2005; 365: 755-763. Pts c stroke/PEG did not do better than those c stroke/NGT.
  • -Sullivan PB, Dev Med Child Neurol 2004; 46: 796-800. gtube improves QOL.

Gastrostomy Tube Review with annotated references: Laparoscopic gtube versus conventional PEG placement

 Zamakhshary et al.  JPS 2005; 40: 859-62.   i.  Retrospective review, n=119 (only 26 with laparoscopy =21%). (2002-2003)  ii.  States same operative time of ~53 min by combining 2nd procedure w PEG (in 77% w PEG).  It takes these authors a long time to perform PEG and gtube change procedures.  Also, it is not noted how many of these 2nd procedures were coordinated with other needed anesthesias.  (Many times a PEG is replaced at the time of another procedure.)   iii.   3 PEG with transcolonic tube, 2 failed PEG –one with peritonitis, 4 with tract disruption when PEG pulled.  Similar rate of local problems (eg granulation tissue).  ARTICLE does not detail when PEG tubes are pulled –VERY high rate of tract disruptions.   iv. Article missing key details regarding size of PEGs & gtube buttons which may impact complications.  v. Cited advantages according to authors:

  1. “eliminates” risk of hollow viscus injury (JH: this is NOT  accurate)
  2. Useful for small infants (<2kg) (JH: usually gtube NOT needed in <2kg)
  3. Enables “ideal” location (JH: this is NOT  accurate)
  4. Primary button ‘advantage’ (JH: DOES NOT cite potential pitfalls like button too tight, possibility of balloon breakdown, possibly higher rate of dislodgment)

 Vervloessem et al. JPS 2009; 18: 93-97.     i. Retrospective review: 1992-2008.  N=467.  ONLY 19 Lap PEG –thus limited ability to provide comparison.  ii. Cites 59 “major complications” due to PEG –Table 2, including “13” new cases of GERD after PEG (or worsened GERD).  Of the major complications, important complications included 1 sepsis death, 7 peritonitis, 5 gastrocolic fistulas, 4 major granulation tissue, and 11 buried bumpers.  iii.  States that VPS is risk factor for infection but does not state whether any Lap gtubes were done in these patients.  iv. Complication rate decreased over the years—p=0.003; thus PEG procedure became safer with time and experience.   Could not demonstrate a decrease in complications with lap gtube versus PEG.  Authors recommend lap PEG in specific situations such as previous abdominal surgery or if not a good puncture site.

 Segal et al. JPGN 2001; 33: 495-500.    i. Retrospective study, n=110 (1990-97). N=110 –ALL PEG (no LAP). Thus, limited utility in comparing two methods. ii. “44%” developed late complications with PEG.  Most common: 24 extruded tubes/buried tubes (would NOT be better with lap button); other important: cologastric fistula n=2, peritonitis.  Table 1 indicates that 75% of dislodgment were due to buttons not PEG.  12 of the complications were granulation tissue and proliferative gastric mucosa.  Buried tubes occurred 14 & 19 months after placement with button tube!!   iii.  Thus this article adds little to the discussion of PEG vs lap gtube.

Akay et al. JPS 2010; 45: 1147-52  i.  Retrospective review (2004-2008) n=238 (134 PEG, 104 LAP)  ii. PEG with higher complications;  authors were changing PEG after 6-8 weeks. iii.  6 patients had early PEG dislodgment –this is higher than expected.  iv. 1 patient with gastrocolic-cutaneous fistula with both PEG & with LAP.  v. Table 4 lists complications: similar stomal issues, 2 patients with leak after PEG exchange (too early! –see page 1152) vi. Cited advantages: “eliminating” risk of hollow viscus injury, allows for sutures, small infants (<2kg) & possible primary buttons.**These authors did not place primary buttons –this makes it difficult to draw any conclusions about PEG vs primary LAP button.  Many feel PEG tube is a better tube and less prone to dislodgment than button and guarantees appropriate size.

Lantz et al. Int J Pediatr 2010; ID# 507616, 1-4.   i. Literature review, included 54 studies that qualified (1995-2009).  N=4331 (1027 LAP, 3304 PEG).  Very few details given in this review.  ii.Fistulas in 1.27% of PEG vs 0% for LAP.  iii.  Lists significant limitations: different studies, not blinded, nonpublication bias.  iv.  “This study highlights the need …for trials, comparing PEG to” LAP. v.  Does not include the limitation that LAP technique developed later and with more experience less complications.  Except for gastrocolic-cutaneous fistulas –no specific information is given about complications.

Avitsland et al.  JPGN 2006; 43: 624-28.  i. Restrospective review. N=121 –all PEGs  ii.     PEG “safe technique…major complications rare.”  “Most children experience minor stoma-related complications.”  iii. 29 died due to other factors.  Of 85 with f/u, 21 able to remove gastrostomy.  iv. No early mortality (<30 days).  1 of 85 had tube dislodgment.  3 had tube migration into esophagus (in cases where tube was not endoscopically removed).  v.     Frequent tube site problems ~75% -most easily treated. vi. Parents with high satisfaction: 83/85 (98%) would choose PEG again, 80/85 (94%) stated PEG improved child’s situation

Gauderer M. JPS 2001; 36: 217-19.   i.  Focused literature search and personal 20 year experience. ii. >216,000 PEGs performed annually in U.S. according to article (~5000 children).  PEG procedure developed 1st for children. iii.  Suggested approach to PEG with or w/o fundoplication: “Because PEG is such a simple procedure, a well-accepted approach is to place gastrostomy initially in children who can tolerate nasogastric tube feedings and add an antireflux procedure later, if needed.

Srinivasan et al. JPGN 2009; 49: 584-88.   i.Prospectively collected data from observational study, n=601 (384 PEG insertions, 165 button conversions).  ALL pediatric. ii.  Complications:  PEG site erythema 15%, buried bumper migration (1 patient), 3 PEG dislodgments, one patient had laparotomy due to severe pain (no findings identified).   No procedure-related mortality.   iii.  49 of 384 removed –no longer needed.  iv. “The role of PEG is well established…our experience..PEG has been generally safe, with low procedure-related morbidity in children.

Nutr Clin Pract 2005; 20 (6): 607-12.  Bankhead RR et al. i. Comparison of 91 patients.  23 PEG, 39 LAP, 29 open.  ii.   PEG had lowest complication rate

Surg Endosc 2006; 20: (8): 1248-51.  Ljungdahl M.                                         i.     Prospective, randomized study. N=70.  ii.  PEG with lower complication rate than surgical (open) gastrostomy –lower mortality & morbidity in adult patients.

UK Review Online:   2009  i. Review of alternatives to PEG for gastrostomy insertion. There are alternative methods of gastrostomy tube insertion to PEG. They are: a) Laparoscopic insertion b) Open surgical technique c)Percutaneous radiologically guided gastrostomy (PRG) insertion. ii.  “There are reports over the years since introduction of PEG in the 1980s with often inconclusive results.21▪    A small study from Ireland and one from London favour PRG in patients with amyotrophic lateral sclerosis as it avoids the need for sedation or endoscopy.22,23▪  One meta-analysis suggested a higher success rate with PRG than with PEG, and less morbidity than either PEG or surgery.24 However a more recent comparison of a relatively small number of endoscopic, surgical and laparoscopic placement favoured PEG25 and another favoured PEG over PRG.26▪     A literature review suggested PEG as the procedure of choice for placement of gastrostomy tubes.27▪    A recent prospective randomized trial favoured PEG over surgical gastrostomy insertion.28▪     There is some evidence that polyurethane PEGs are less troublesome than silicone PEGs (less tube deterioration, less blockage).29▪    PEG is preferred in trauma patients.30▪                Antibiotic prophylaxis for PEG insertion appears to reduce the incidence of wound infection.19,20▪      Laparoscopic insertion was considered preferable to PEG by one study in children with PEG insertion having higher complication rate in children and often requiring repeat anaesthetics.31   An earlier study in children showed similar results for surgical, PRG and PEG insertion but did not look at the laparoscopic technique.32    A recent study from Norway found PEG insertion safe and very well tolerated by children and parents but made no comparison with other techniques.

Conclusions of review: PEG likely increases risk of gastrocolic fistulas (1-2%) but this has been reported with LAP as well.  The incidence is low.  No well-designed  studies have demonstrated superiority of LAP over PEG in terms of safety.  Potential drawbacks of LAP are likely underreported.  There have been cases of severe peritonitis at local hospitals following lap with primary gtube balloon misplacement.  Many feel PEG tube is a better tube and less prone to dislodgment than button (dislodgment is most frequent serious adverse event) and can be easily adjusted to  appropriate size.  To minimize complications, tube should not  be changed early.