In our experience LRFJ was successful in 11/12 patients with delayed gastric emptying. Although complications occurred in 8/12 patients most of them were minor complications and parents were satisfied in 11/12 cases.
LRFJ is a rare surgical procedure and to our knowledge this is the second study reporting on LRFJ in the pediatric population. This makes it difficult to compare our experiences with others. Our study shows a similar amount of complications compared to the first report on LRFJ in children by Neuman et al. who reported about LRFJ in five patients . Only minor complications and a delayed start of jejunal feeding in two patients were found. The delayed start of feeding was due to emesis and pain related to visceral hyperalgesia and idiopathic diarrhea. These outcomes correspond with our experiences.
More publications about open Roux-en-Y jejunostomy are available. Recently a systematic review comparing these publications was published . Open surgery is associated with a different type (more severe), and a higher rate of complications compared to the LRFJ, as there is a high (6-40%) incidence of wound infections [17, 23, 24] and volvulus (14-25%) [24-26] compared to respectively 8% and 0% in our patient population. Despite this difference the open jejunostomy showed, as expected, the same complications related to the jejunostomy site as the laparoscopic jejunostomy site: leakage (17-43%) [4, 17, 23] and hypergranulation (no percentages available) . In our patient population leakage occurred in 17% of the patients. In another study with 11 patients only long-term outcomes but no complications were described .
The incidence of volvulus is probably related to the length of the Roux limb and the kind of surgery (open or laparoscopic). Taylor et al. reported, in a series of 25 open Roux-en-Y jejunostomy procedures, a small bowel volvulus around the Roux limb in 20% of the patients. The patients with a volvulus showed a relatively longer Roux limb compared to the patients without volvulus (18.7 +- 7.7 vs 14 +- 2.3 cm). In one patient with a volvulus the Roux limb was only 6 cm . In the studies of Singh et al. and McCann et al. no details about Roux limb length were described but all patients with volvulus received open surgery [24,26]. In our series the Roux limb was relatively short (10-15 cm). No volvulus or stenosis occurred until now.
It is interesting to see that nearly all patients in our study started jejunal feeding one day postoperatively while patients described in other studies started jejunal feeding around 3-7 days after surgery [19, 23, 28]. No clear explanation was given for this delay. Only one study described good results with early (<48h) jejunal feeding in 13 patients just like we found .
Strengths and limitations
Despite good results our study has its limitations. This is because of the retrospective study nature and the relative small number of patients. The retrospective nature makes it for example difficult to know for certain if every complication is noted. Besides this our study is a case-series with little to no comparison which makes it difficult to compare outcomes to other studies.
In the literature no data was provided on parental satisfaction and we could not compare our results to parental satisfaction after open Roux-en-Y jejunostomy/LRFJ in other surgical centers.
In addition the diagnosis delayed gastric emptying was not clearly defined in our population and therefore the decision to perform LRFJ was mainly based on clinical reasons after ruling out other conditions. The role of gastric emptying scans/studies was supportive as the outcomes of the gastric emptying scans/studies differed and did not influence the decision to perform LRFJ on these patients.