ABSTRACT
Upper Gastro-Intestinal Bleeding (UGIB) is one of the common complaints with which patients present to casualty. It is associated with significant morbidity and mortality. The aetiological spectrum of UGIB is variable in different geographical regions. Our study aimed to analyse the aetiology, endoscopic profile, mortality, Rockall score and predictors of mortality in patients with UGIB, in North East India. METHODSThis cross-sectional study was conducted at Assam Medical College and Hospital in North East India. We enrolled patients with age 12 years and above, who were admitted between July 2019 and January 2020 with a history suggestive of UGIB. Demographic data of the patients was collected, after which they underwent clinical examination, and upper GI endoscopy. Mean ± standard deviation was used to express continuous variables. Frequency and percentage were used to express categorical variables. Test of significance for qualitative data was assessed by Chi-square test (for 2 x 2 tables). P value less than 0.05 was taken as statistically significant. RESULTSWe analysed 117 patients diagnosed with UGIB [80.34 % male, 19.60 % female], ratio of male to female of [4.08:1] was seen. The most common symptom was melena 87 patients (74.15 %), endoscopy finding showed that 48.71 % had oesophageal and / or gastric varices, 26.49 % had peptic ulcers, 17.94 % had gastric erosions / duodenal erosions / erosive gastritis, 1.7 % had Mallory-Weiss tear, 1.7 % had gastric malignancy, 1.7 % had GJ stoma bleed, 1.7 % had both oesophageal varices and peptic ulcer disease. Partial gastric outlet obstruction was observed in peptic ulcer disease in 2 patients (6.45 % of total peptic ulcer disease patients). 73.75 % patients had Rockall score < 5 and 26.49 % patients had Rockall score > 6. H. pylori infection (assessed by RUT) was an independent predictor of upper GI bleed in both variceal and non-variceal bleed [p < 0.001]. The mortality in our study was 7.69 %. Predictors of mortality in the study population were, patients with variceal bleed [p = < 0.001], Rockall score > 6 [p = 0.013], and chronic liver disease [p < 0.001]. The average duration of hospital admission of the study population is about 4.6 + / - 0.4 days. CONCLUSIONSThe study reported oesophageal varices was the most common cause of UGIB, followed by peptic ulcer in North East India. H. pylori was an independent predictor of both variceal and non-variceal bleed. Partial gastric outlet obstruction (GOO) was one of the common benign complication of peptic ulcer disease. Variceal bleed, Rockall score > 6, chronic liver disease were predictors of mortality.