Ascites predicts the presence of high grade varices by screening gastroscopy

J. Lavergne, E. Molina, K. R. Reddy, L. Jeffers, R. Leon, A. K. Nader, E. R. Schiff

Research output: Contribution to journalArticlepeer-review

18 Scopus citations


Screening gastroscopy is recommended in cirrhotic patients to rule out large esophageal and/or gastrk: varices. When grade III or IV esophageal varices are present, primary prophylaxis with beta blockers is indicated. As varices are a manifestation of portal hypertension, we wanted to determine if we could predict the endoscopic presence of large varices (grade III or higher) based on other manifestations of portal hypertension, Child Pugh classification, or other independent variables. Materials: The charts of the 52 patients (pts) who had screening EGDs to rule out varices were retrospectively reviewed. Data was collected for age, sex, white blood cell count (wbc), hemoglobin (Hb), platelet count (Pit), albumin, bilirubin, prothrombin time, porto-systemic encephalopathy (PSE) requiring treatment, the presence of ascites, spler omegaly , and/or intraabdominal varices by imaging studies (US/CT). Results:The mean age was 66, with 31 males and 21 females. Eleven (48.8%) patients were Child-Pugh class A, 16 (35.5%) class B, and 7(15.5%) class C. The mean values of the collected data with their 95% confidence intervals were: wbc 4.85 (4.24-5.45), Hb12.3 (11.7-12.8) g/dl, Pit count 92.8 (75.6-110.1), bilirubin 2.12 (1.45-2.78)mg/dl, albumin 3.58 (2.99-4.18)g/dl, and PT 14.5 (13.9-15.1) seconds. A total of 45 pts had imaging studies: ascites was present in 14 (31%)pts, splenomegaly in 35(77%) pts, and intraabdominal varices by US/CT in 12 (26.6%). PSE was recorded in 10 (19.2%). Endoscopic esophageal varices (EEV) were present in 41 of 52 pts (78.8%): grade 1 in 15pts (28%), grade II in 16pts (30.7%), grade III in 8 (15%), and grade IV in 2 (3.8%). Six of 14 (42.8%) patients with ascites had grade III/IV esophageal varices versus 2 of 31 (6.45%) without ascites (p=0.007 by Fisher exact two tail and p=0.0: 12 by Yates method). Except for ascites, none of the other mentioned variables had a significant p value by univariate analysis when correlated to the presence or absence of EEV (grade I and II versus III and IV). Conclusions: Ascites significantly correlates with the presence of varices that require primary prophylaxis. By univariate analysis Child-Pugh class, clinical parameters, imaging or laboratory tests did not predict the presence of high grade varices.

Original languageEnglish (US)
Pages (from-to)AB187
JournalGastrointestinal endoscopy
Issue number4
StatePublished - Jan 1 1997

ASJC Scopus subject areas

  • Radiology Nuclear Medicine and imaging
  • Gastroenterology


Dive into the research topics of 'Ascites predicts the presence of high grade varices by screening gastroscopy'. Together they form a unique fingerprint.

Cite this