Abstract Background A feeding jejunostomy is one method to provide adequate nutrition after oesophagectomy. This bypasses the anastomosis allowing healing before oral feeding resumes. However, routinely using feeding jejunostomies may not be appropriate for patients within an Enhanced Recovery After Surgery protocol due to complications. Feeding jejunostomies have 3–61% morbidity, with mechanical (i.e tube dislodgement) and non-mechanical (i.e. site infection) complications plus associated increased length of stay. The local standard after oesophagectomy is routine feeding jejunostomy placement. However, selective, rather than routine feeding jejunostomy placement may therefore be warranted. Aim Conduct a quality improvement project addressing the local feeding jejunostomy complication rate after oesophagectomy. Methods Cycle 1: A review of oesophagectomies between 2018–2021 assessing complications using the Clavien-Dindo classification. An existing database of oesophagectomies was screened for patients undergoing concurrent feeding jejunostomy insertion. Patients ≥18 were included, whilst any indication for oesophagectomy was included. Patients lacking information on type and bore of jejunostomy tube inserted were excluded. Demographics, jejunstomy tube type and bore, complication and Clavien-Dindo grade of complication were recorded. The Charlson comorbidity score of each patient was determined to help find a population at risk. Intervention: Findings to be presented at a departmental audit meeting, with patient risk stratification to identify a cohort at risk. Cycle 2: re-audit (July 2022 - July 2023) as per cycle 1 after presentation. Improvement = fewer jejunostomy-specific complications. Results 110 patients were included, with a median age of 65.5 (IQR 61.0–72.0). 81.8% patients (n=90) sustained a complication, mostly Clavien-Dindo grade II. Jejunostomy-specific complications occurred in 21.8% patients (n=24), with the commonest jejunostomy-specific complication being a leak (44.8%). 4 ‘major’ (Clavien-Dindo ≥III) jejunostomy-specific complications were reported. No ischaemic small bowel necrosis was reported, significantly less than a prior audit (0.0% vs. 3.74%, p=0.029). Demographics/ Charlson comorbidity score, jejunostomy type inserted (JJ tube, 15F or 9F) and jejunostomy insertion method (open/ laparoscopic) did not impact overall or major feeding jejunostomy-specific complication rate. Conclusions This first cycle agrees with the reported complication rate in the literature. Though a population at risk could not be determined, reviewing local non-specialist centres may aid this, to reduce future complications. The literature would support a more selective approach to feeding jejunostomy insertion and as such this should be taken to reduce post-insertion complications. Continued prominent advice regarding inotropes may prevent ischaemic bowel. We aim to introduce clear warnings to the local ERAS protocol over the concurrent use of jejunostomy feed and inotropes, whilst electronic warnings flagging concurrent jejunostomy feed and inotrope prescribing are being worked upon. Moreover, poor hospital coding of jejunostomies may have limited the cohort, preventing determining a cohort at risk of feeding jejunostomy insertion.
Read full abstract