Non invasive predictors for presence of esophageal varices in patient of liver cirrhosis-a cross sectional analytic study

Objective: Objective of this study was to identify non invasive predictors for the presence of Esophageal varices in patients of liver cirrhosis without prior history of upper Gastrointestinal bleeding. Methodology: This cross sectional analytic study was conducted from August 2011 to July 2016 in the Department of Gastroenterology at Liaquat National Hospital Karachi and Aga Khan University Hospital, Karachi. Patients diagnosed to have liver cirrhosis without prior history of upper GI (gastrointestinal) bleeding secondary to varices either visiting the outpatient department or admitted in hospital were included. Informed consent was taken from all patients. Screening Esophagogastroduodenoscopy (EGD) was performed in all patients to detect varices once the diagnosis of liver cirrhosis was confirmed. Patients of both gender and age more than 16 years were included in the study. Statistical package of social science (SPSS) were used to analyze data. Results: Hundred patients of liver cirrhosis who fulfilled the selection criteria were included in the study. Average age of the patients was found to be 48.71+12.38, out of 100 patients 46% were male and 56% were female. Esophageal varices were seen in 76/100 (76%) patients, significant large varices (Grade III and IV) seen only in 9.2 % of patients. 37(47.7%) Patients with Varices were male and 39(51.3%) were female. Univariate and multivariate analysis of variables was performed between patients with and without varices. Serum albumin, bilirubin, portal vein diameter were significantly associated with presence of esophageal varices whereas platelets counts, prothrombin time and splenic size were not significant in univaiate analysis. Multivariate analysis show low platelets, decreased albumin, increased prothrombin time, increased serum bilirubin and increased portal vein diameter were significant independent predictors of presence of esophageal varices. Conclusion:Incidence of esophageal varices in patients with liver cirrhosis without a history of upper GI bleeding was significantly high our study. The statistically substantial predictors of presence of esophageal varices were, platelets count<150,000/cmm, serum albumin <2.8gms /dl, serum bilirubin>1.2mg /dl , prothrombin time >2 seconds of control value , portal vein diameter >11mm.