[Risk Factors Associated with Failure from Endoscopic Therapy in Acute Non-Variceal Upper Gastrointestinal Bleeding].
Jia-ying Zhang,Ye Wang,Jing Zhang,Shi-gang Ding,Li-ya Zhou,San-ren Lin
DOI: https://doi.org/10.3969/j.issn.1671-167x.2010.06.022
2010-01-01
Abstract:OBJECTIVETo determine risk factors associated with failure of endoscopic therapy in acute non-variceal upper gastrointestinal bleeding (ANVUGIB ).METHODSThis was a retrospective cohort study of 223 patients admitted to Peking University Third Hospital between 1 January 2005 and 31 December 2009, with acute non-variceal upper gastrointestinal bleeding. Data on clinical presentation, laboratory test, endoscopic findings, and treatment outcomes were collected. Risk factors for treatment failure were identified using multivariable Logistic regression with backward selection.RESULTSTherapeutic failure rate was 19.3%(43/223). In univariate analysis, the two groups had significant difference in age, history of gastrointestinal bleeding, ASA, shock, haemoglobin level, Hct, PLT, time of endoscopic treatment, gastric ulcer, duodenal ulcer, lesion size and active spurting of blood. Multivariate Logistic regression analysis revealed that shock [odds ratio (OR) 3.058, 95% confidence interval (CI) 1.295-7.221], history of gastrointestinal bleeding (OR 2.809, 95% CI 1.207-6.539), PLT>100×10⁹/L (OR 0.067, 95% CI 0.009-0.497), active spurting of blood (OR 10.390, 95% CI 2.835-38.080) and lesion size≥2.0 cm (OR 7.111, 95% CI 1.628-31.069) were risk factors associated with failure of endoscopic therapy. The number of comorbidities>1 (OR 9.580,95%CI 1.383-66.390) and active spurting of blood (OR 9.971, 95% CI 1.820-54.621) were factors related with need for surgical intervention or death.CONCLUSIONPatients with shock, history of gastrointestinal bleeding, PLT<100×10⁹/L, active spurting of blood and large lesion size, have high risks for continued bleeding or rebleeding after endoscopic treatment. These patients may be more likely to benefit from aggressive post-hemostasis care.