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


Review: β-blockers and sclerotherapy prevent first bleeding from varices

ACP J Club. 1992 Nov-Dec;117:66. doi:10.7326/ACPJC-1992-117-3-066

Source Citation

Pagliaro L, D'Amico G, Sørensen TI, et al. Prevention of first bleeding in cirrhosis. A meta-analysis of randomized trials of nonsurgical treatment. Ann Intern Med. 1992 Jul 1;117:59-70.



To assess the effectiveness of β-blockers and endoscopic sclerotherapy in the prevention of first bleeding and mortality in patients with cirrhosis and esophagogastric varices.

Data sources

Studies were identified through MEDLINE searching (1980 to 1990), author contacts, bibliography review of relevant studies and review articles, and conference abstract lists.

Study selection

Studies enrolling patients with cirrhosis and esophageal varices that had never bled were retrieved. Trials were included only if mortality and the incidence of first bleeding were reported. 9 randomized trials (7 full articles, 2 abstracts) of β-blockers compared with nonactive treatment and 19 trials (15 articles, 4 abstracts) of sclerotherapy compared with nonactive treatment were analyzed.

Data extraction

The validity of the published trials was evaluated by 3 independent investigators using predetermined criteria. Bleeding and mortality rates were the outcome measures evaluated. Common and pooled odds ratios and the number needed to treat (NNT) to prevent an outcome were calculated. Heterogeneity was examined using Peto-Mantel-Haenszel methods and L'Abbé plots.

Main results

The 9 trials of β-blockers included 996 patients. 2 trials were double blind and 4 were single blind. The heterogeneity of the β-blocker trial results evaluating first bleeding (P = 0.007) was attributed to a single outlier study (odds ratio [OR] 4.90, 95% CI 1.23 to 19.5). The pooled OR of all 9 trials examining first bleeding was 0.54 (CI 0.39 to 0.74). For mortality the pooled OR was 0.75 (CI 0.57 to 1.06). The number of patients who were treated with β-blockers to prevent 1 additional episode of bleeding was 11 (CI 8 to 18). Of 19 sclerotherapy trials evaluating 1630 patients, only 1 study was single blind. The pooled OR for first bleeding was 0.6 (CI 0.49 to 0.74); for mortality it was 0.76 (CI 0.61 to 0.94). The heterogeneity of study results was significant (P < 0.001 for both first bleeding and mortality). Explanations included a greater apparent treatment effect in patients at higher baseline risk for bleeding and a greater treatment effect in studies of lower methodologic quality.


β-blockers and sclerotherapy prevent first bleeding in patients with cirrhosis and high-risk varices. The effect of these 2 interventions on mortality is similar; the benefit, however, reaches conventional levels of statistical significance only for sclerotherapy.

Source of funding: Not stated.

For article reprint: Dr. G. D'Amico, Clinica Medica R, Divisione di Medicina Generale, Ospedale V Cervello, Via Trabucco 180, 90146, Palermo, Italy.


Scientific overviews that include a comprehensive search for relevant literature, an unbiased assessment of the validity of the primary research, and an examination of the reasons for differences in study results can provide important insights into treatment effects (1). The meta-analysis by Pagliaro and colleagues provides compelling evidence for prophylactic β-blockers or sclerotherapy for primary prevention of gastrointestinal bleeding in patients with cirrhosis and esophagogastric varices. Bleeding was reduced by β-blockers although the reduction in mortality was not statistically significant. Sclerotherapy significantly reduced bleeding rates and mortality, although the trial results were heterogeneous.

Heterogeneity of trial results warrants closer examination of the data. In the sclerotherapy trials, variability in results was explained by different risks for bleeding across studies and larger treatment effects in trials of lower methodologic quality. The best estimate of the efficacy of this intervention, nevertheless, is the pooled odds ratio, which suggests substantial risk reductions.

Before all clinicians endorse prophylactic sclerotherapy, however, further studies are needed to identify clinical predictors of bleeding (2) and subgroups of patients likely to derive the most benefit. Prophylactic administration of β-blockers for patients with cirrhosis and esophagogastric varices should be considered for all patients without contraindications.

Marcos Pedrosa, MD
Department of Veterans AffairsMedical CenterBoston, Massachusetts, USA
Deborah J. Cook, MD
McMaster UniversityHamilton, Ontario, Canada

Marcos Pedrosa, MD
Department of Veterans AffairsMedical Center
Boston, Massachusetts, USA

Deborah J. Cook, MD
McMaster University
Hamilton, Ontario, Canada


1. Bossel JP, Blanchard J, Panak E, Peyrieuz JC, Sacks H. Controlled Clin Trials. 1989;10:254-81.

2. The North Italian Endoscopic Club for the Study and Treatment of Esophageal Varices. N Engl J Med. 1988;319:983-9.