Endoscopic Screening for Varices in Cirrhotics: Data From a National Endoscopic Database (the CORI Project)

Abstract

Endoscopic Screening for Varices in Cirrhotics: Data From a National Endoscopic Database (the CORI Project) Mary Kovalak, Jeremy Lake, Nora Mattek, David Lieberman, Glenn Eisen, Atif Zaman Purpose: The current American College of Gastroenterology (ACG) guidelines for primary prophylaxis of variceal hemorrhage in cirrhotics recommend screening for varices with upper endoscopy (EGD). The national screening pattern of endoscopists is not known. Patient characteristics associated with the finding of varices on screening endoscopy are not fully established. Aim: Evaluate national endoscopic practice patterns and factors associated with finding varices on screening endoscopy using a national endoscopic database. Methods: Endoscopic reports generated by the Clinical Outcomes Research Initiative (CORI) repository were analyzed to determine the annual variation and findings of screening upper endoscopies between 1/1/2000 and 12/31/2003. Results: Data from 69 CORI practices (44 community/16 academic/9 Veterans Affairs Medical Center (VAMC)) was evaluated. A total of 1688 of 172,854 EGDs (9.8 per 1,000 EGDs) were performed for the purpose of screening for varices. Overall, there was a linear increase in annual proportion of screening EGDs performed. A significantly greater proportion of variceal screening EGDs are performed in academic centers (28.3 per 1,000 EGDs) compared to VAMC (10.8 per 1,000 EGDs) and community (5.3 per 1,000 EGDs). Among screening EGDs (n Z 1688), 881 (52.2%) yielded varices. Varices were found more often in Child-Pugh class B/C (71.9%) compared to ChildPugh class A (42.7%) patients. Of those with varices found, patients with Child-Pugh class B/C were more likely to have large varices (vs. small/med sized) than patients with Child-Pugh class A (p Z 0.02). Among procedures with an etiology reported (78%), patients with alcohol as an etiology of cirrhosis were significantly more likely (!.001) to have varices found than those who did not. Conclusions: The data shows 1) The proportion of EGDs performed for variceal screening has increased between 2000 and 2003. 2) A higher proportion of EGDs are performed for variceal screening in academic centers compared to VAMC and community. 3) Varices were found more frequently in Child-Pugh classes B and C and in patients with cirrhosis due to alcohol. 4) Patients with Child-Pugh class B/C were more likely to have large varices than patients with Child-Pugh class A. Abstracts