Background: Esophageal variceal bleeding is a potentially fatal complication in patients with liver cirrhosis and portal hypertension. In cirrhotic patients, endoscopic screening for esophageal varices (EV)is currently recommended at the time of diagnosis. The present study intends to find out correlation between platelet countspleen bipolar diameter ratio and esophageal varices in liver cirrhosis and prospectively validate its use for the noninvasive diagnosis of EV.Methods: This observational study was done at theDepartment of Gastrointestinal Hepatobiliary and Pancreatic Disorder, BIRDEM General Hospital, Dhaka, during the period of November, 2013 to October, 2014. A total 64 patients with cirrhosis of liver were included. Complete blood count, liver function tests, ultrasonographyof whole abdomenand endoscopy of upper gastro-intestinal tract (GIT)were done in all patients. Statistical analysis was done with SPSS version 16.Results: Among 64 study population EV were detected in 54 patients (84.4%). The platelet count/ spleen bipolar diameter ratio was significantly higher without EV compared with those with EV(1570 ± 493) and (688 ± 227), respectively; (p<0.001).Patients with EV had lower platelet counts (86799.84 ±27389.99/mm3, p<0.001), higher bipolar spleen diameters (127.94±15.14 mm, p<0.001) and lower platelet count/ spleen bipolar diameter ratios (688.79 ±227.13, p<0.001). In this analysis, by applying (ROC)curve the platelet count-spleen bipolar diameter ratio at a cut-of 908.5 maintained high sensitivity (100 %), Robust negative predictive value NPV (100%),specificity (55.6%) and PPV (85.4%), with an overall diagnostic accuracy of 87.50% for the prediction of EV.With advancing Child-Pugh (C-P) class the percentage of patients with varices increased, 33.3% in C-P class A, 85.7% in C-P class B and 91.6% in C-P class C. The presence of EVcorrelated significantly with the severity of liver cirrhosis (p = 0.03) as measured by Child-Pugh score.Conclusion: The identification of non-endoscopic, noninvasive methods that can accurately predict esophageal varicesin cirrhosis of liver has been addressed in several recent studies. This study was yet another attempt to achieve this goal. Predicting the presence of EV by noninvasive means would restrict the performance of endoscopy and reducing the number of unpleasant screening endoscopiesBirdem Med J 2018; 8(2): 159-166

Objective: To determine etiological pattern of patients presenting with upper gastrointestinal bleeding in a tertiary care hospital of Bangladesh.Methods: This study was a prospective observational study, carried out in the Department of Gastroenterology, Bangabandhu Sheikh Mujib Medical University (BSMMU), Dhaka, Bangladesh from January 2013 to June 2013. Fifty adult patients presenting with haematemesis and/or melaena admitted into gastroenterology inpatient unit from outpatient department or patients referred from other inpatient units of Department of Medicine of BSMMU were included in the study. Endoscopic examination was performed within 24 to 48 hours of presentation. Lower GI endoscopy was done in selected cases.Results: The study patients were predominantly young or of early middle age with mean age being 34.45 ± 16.5 yrs. A male preponderance was observed with male to female ratio being 7.3:1 (44 male and 6 female). 62% of the patients presented with both haematemesis and melaena, 26% with melaena only and 12% with haematemesis alone. Endoscopy of upper gastrointestinal tract demonstrated duodenal ulcer to be predominant finding (50%), followed by gastro-esophageal erosions (20%), gastric ulcer (12%), esophageal varices (10%), gastric adenocarcinoma (4%) and stomal ulcer (4%).Conclusions: Endoscopy revealed that duodenal ulcer was the most common cause of upper gastrointestinal bleeding followed by oesophageal erosion, gastric ulcer, esophageal varices and stomal ulcer. Peptic ulcer disease still remains as the major cause of acute upper gastrointestinal haemorrhage, though cases of oesophageal erosion were also significant.Bangladesh Crit Care J September 2015; 3 (2): 60-62