Introduction There is an emerging group of children in whom poor and worsening upper GI dysmotility limits feed toleration and impacts growth; we wished to evaluate the role of jejunal tube feeding (JTF) in this group.

Methods A retrospective cohort study (database/clinical note review) in a tertiary paediatric centre to evaluate use of PEG-J, transgastric gastrojejunostomy (GJ) tubes and surgical roux-en Y jejunostomy (ReYJ), and the impact on growth of JTF in children with worsening GI dysmotility. All children (<18 years) receiving home enteral tube feeding (HETF) during the period 01 January 2002–31 December 2011. Weight at time of commencing JTF and at 6 or 12 months post-start was collected and expressed as SD or Z-score. Change in weight Z-score was calculated using paired t-test.

Results A total of 866 children received HETF during the study period, of whom 41(5%) had JTF at home. Median (range) decimal age at start of JTF was 2.7(0.1–16.2) years. 36 of 41 (88%) had an underlying neurodisability; 33 of 41 (80%) were gastrostomy fed prior to commencing JTF. Of the 41 JTF children, 19 (46%) were fed via a GJ tube, 5 (12%) via PEG-J and 17 (42%) had a ReYJ. The majority of JTF related complications occurred with GJ tubes; although usually minor, one death occurred following small bowel intussusception around a GJ tube. Minor JTF complications included burst balloons, holes in the Y-port or tube and fungal infection and resolution required tube changes. Tube migration was a problem with both GJ and PEG-J tubes; ReYJ were associated with the fewest minor complications of stomal infection and leakage. By study end, 21 (51%) continue on JTF, 9 (22%) died (all but 1 due to their underlying condition), 1 (2%) moved out of area, 2 (5%) transitioned to adult services and 8 (20%) returned to gastric feeding. 25 of 41 children had JTF for >6 months and had longitudinal growth data collected; median (range) weight Z-score at the start of JTF was −1.3 (−5.2–2.1) and rose to −1.0 (−3.4–2.3) by 6–12 months, with a significant improvement in mean (95% CI) change in weight Z-score of 0.7 (0.1 to 1.3) (p=0.02).

Conclusion There are time consuming practical challenges associated with JTF, some of which are device dependent, and ReY JTF appears best for long-term usage. JTF is an effective intervention to improve growth in children with severe and worsening upper GI dysmotility.

Competing interests None declared.