Selected nasogastric lavage in patients with nonvariceal upper gastrointestinal bleeding

BACKGROUND: Risk stratification before endoscopy is crucial for proper management of patients suspected as having upper gastrointestinal bleeding (UGIB). There is no consensus regarding the role of nasogastric lavage for risk stratification. In this study, we investigated the usefulness of nasogastric lavage to identify patients with UGIB requiring endoscopic examination. METHODS: From January 2017 to December 2018, patients who visited the emergency department with a clinical suspicion of UGIB and who underwent nasogastric lavage before endoscopy were eligible. Patients with esophagogastric variceal bleeding were excluded. The added predictive ability of nasogastric lavage to the Glasgow–Blatchford score (GBS) was estimated using category-free net reclassification improvement and integrated discrimination improvement. RESULTS: Data for 487 patients with nonvariceal UGIB were analyzed. The nasogastric aspirate was bloody in 67 patients (13.8 %), coffee-ground in 227 patients (46.6 %), and clear in 193 patients (39.6 %). The gross appearance of the nasogastric aspirate was associated with the presence of UGIB. Model comparisons showed that addition of nasogastric lavage findings to the GBS improved the performance of the model to predict the presence of UGIB. Subgroup analysis showed that nasogastric lavage improved the performance of the prediction model in patients with the GBS ≤ 11, whereas no additive value was found when the GBS was greater than 11. CONCLUSIONS: Nasogastric lavage is useful for predicting the presence of UGIB in a subgroup of patients, while its clinical utility is limited in high-risk patients with a GBS of 12 or more. SUPPLEMENTARY INFORMATION: The online version contains supplementary material available at 10.1186/s12876-021-01690-z.