Current issues of ACP Journal Club are published in Annals of Internal Medicine


Review: Ligation is the best endoscopic choice for esophageal variceal bleeding

ACP J Club. 1996 Jan-Feb;124:11. doi:10.7326/ACPJC-1996-124-1-011

Source Citation

Laine L, Cook D. Endoscopic ligation compared with sclerotherapy for treatment of esophageal variceal bleeding. A meta-analysis. Ann Intern Med. 1995 Aug 15;123:280-7. [PubMed ID: 7611595]



To compare endoscopic ligation with sclerotherapy for treating esophageal variceal bleeding.

Data Sources

Studies were identified through MEDLINE using the search terms varices, ligation, and band; BIOSIS Previews; Federal Research in Progress database; citation searches of relevant articles; lists and databases of conference proceedings and abstracts; bibliographies of relevant papers; and contact with authors and manufacturers.

Study Selection

Studies were selected if they were randomized controlled trials comparing endoscopic ligation with sclerotherapy; if the target population included patients with esophageal variceal bleeding; and if outcomes measured were rebleeding, mortality, complications, or the number of treatment sessions to obliteration.

Data Extraction

Data were extracted on study quality, study population, intervention, disease severity and type (active bleeding, alcoholic cirrhosis, and Child C liver disease), and outcomes.

Main Results

158 studies were identified, and 7 (4 full publications and 3 abstracts) met the inclusion criteria. 547 patients were included (mean age range 46 to 56 y). No evidence of heterogeneity was found; data were pooled. Endoscopic ligation compared with sclerotherapy had lower rates of rebleeding {31% vs 46%, 95% CI for the 15% absolute risk reduction [ARR] 7% to 23%, P < 0.001; relative risk reduction [RRR] 33%, CI 17% to 46%; number needed to treat [NNT] 7, CI 4 to 14}*, rebleeding caused by varices {21% vs 36%, CI for the 16% ARR 6% to 26%, P = 0.004; RRR 42%, CI 16% to 60%; NNT 7, CI 4 to 17}*, mortality {24% vs 32%, CI for the 8% ARR 0.1% to 15%, P = 0.05; RRR 25%, CI 1% to 42%; NNT 13, CI 7 to 1000}*, mortality caused by bleeding {7% vs 14%, CI for the 7% ARR 0.1% to 12%, P = 0.05; RRR 47%, CI 0 to 72%; NNT 15, CI 8 to 1000}*, and the complication of esophageal stricture {0% vs 10.9%, P < 0.001}*. Overall, 2.2 (CI 0.9 to 3.5) fewer endoscopic treatment sessions were required with ligation to achieve variceal obliteration. Endoscopic ligation did not differ from sclerotherapy for variceal obliteration, hemostasis for active bleeding or the complications of bleeding caused by treatment-induced ulcerations, pulmonary infection, bacterial peritonitis, and complications leading to death. Subgroup analyses (methodology scores, proportion of patients with alcoholic cirrhosis or Child C liver disease, or publication status) showed no differences for rebleeding and mortality.


Compared with sclerotherapy, endoscopic ligation has lower rates of rebleeding, mortality, and complications and requires fewer treatments for obliteration of esophageal varices.

Source of funding: Not stated.

For article reprint: Dr. L. Laine, Gastroenterology Division (LAC 12-137), Department of Medicine, University of Southern California School of Medicine, 2025 Zonal Avenue, Los Angeles, CA 90033, USA. FAX 213-226-7573.

*Numbers calculated from data in article.


Meta-analysis is a highly structured, quantitative approach to reviewing the medical literature that greatly benefits the busy practitioner. By statistically compiling information from several studies, meta-analysis can help make sense of a confusing literature. The results are especially valuable when individual studies are too small to reach statistical significance or when sample sizes are insufficient to support subgroup analyses. The article by Laine and Cook admirably shows the best aspects of meta-analysis.

Laine and Cook comprehensively retrieved and carefully reviewed 158 potentially relevant studies and found 7 that met rigorous and reproducible quality standards. When the results were pooled, ligation therapy was found to be better than sclerotherapy on almost all counts. Patients who had ligation were less likely to rebleed, die, or develop esophageal strictures. Serious complications, such as pulmonary infection, peritonitis, and bleeding from treatment-induced ulcerations, were substantially less common in patients in the ligation group; however, these results did not reach conventional levels of statistical significance and were based on unblinded observations. Subgroup analysis showed that ligation therapy was equally effective for patients with alcoholic cirrhosis or advanced liver disease, and this was so whether studies were published in full form or as abstracts.

Surprisingly, Laine and Cook argue for further large-scale studies to compare ligation with sclerotherapy. Some might argue that their meta-analysis makes additional studies difficult to justify. The evidence indicates that ligation is superior to sclerotherapy for bleeding esophageal varices.

Robert S. Sandler, MD, MPH
University of North Carolina at Chapel HillChapel Hill, North Carolina, USA