Abstract
Background Early oral intake (EOI: initiated within 1 day) and early nasogastric tube removal (ENR: removed <= 2 days) post-esophagectomy is controversial and subject to significant variation. Aim Our aim is to provide the most up-to-date evidence from published randomized controlled trials (RCTs) addressing both topics. Methods We searched MEDLINE and Embase (1946-06/2019) for RCTs that investigated the effect of EOI and/or ENR post-esophagectomy with gastric conduit for reconstruction. Our main outcomes of interest were anastomotic leak, aspiration pneumonia, mortality, and length of hospital stay (LOS). Pooled mean differences (MD) and risk ratios (RR) estimates were obtained using a DerSimonian random effects model. Results Two reviewers screened 613 abstracts and identified 6 RCTs eligible for inclusion; 2 regarding EOI and 4 for ENR. For EOI (2 studies, n = 389), was not associated with differences in risk of: anastomotic leak (RR: 1.01; 95% CI: 0.407, 2.500; I-2: 0%), aspiration pneumonia (RR: 1.018; 95% CI: 0.407, 2.500), mortality (RR: 1.00; 95% CI: 0.020, 50.0). The LOS was significantly shorter in the EOI group: LOS (MD: -2.509; 95% CI: -3.489, -1.529; I-2: 90.44%). For ENR (4 studies, n = 295), ENR (removed at POD0-2 vs. 5-8 days) was not associated with differences in risk of: anastomotic leak (RR: 1.11; 95% CI 0.336, 3.697; I-2: 25.75%) and pneumonia group (RR: 1.11; 95% CI: 0.336, 3.697; I-2: 25.75%), mortality (RR: 0.87; 95% CI: 0.328, 2.308; I-2: 0%)or LOS (MD: 1.618; 95% CI: -1.447, 4.683; I-2: 73.03%). Conclusions Our analysis showed that EOI as well as ENR post-esophagectomy do not significantly increase the risk of anastomotic leak, pneumonia, and mortality. The LOS was significantly shorter in the EOI group, and there was no significant difference in the ENR group. A paucity of RCTs has evaluated this question, highlighting the need for further high-quality evidence to address these vital aspects to post-esophagectomy care. Systematic review registration CRD42019138600