Selection of patients and operations for bleeding esophageal varices
Portosystemic shunts were first introduced in 1945.1 That a high level of interest in this subject persists is attested to by the more than 300 reports that have been published in the past few years. This interest has developed because of high morbidity and mortality in cirrhotic patients with bleeding esophageal varices. Some studies have indicated that less than one third of patients who sustain a major variceal hemorrhage survive 1 or more years after bleeding.2
The theoretical objective of portacaval shunts is reduction of the incidence of variceal hemorrhage and extension of long-term survival of cirrhotic patients beyond that of a nonoperated group. Presumably these operations should be accompanied by minimal morbidity, mortality, and metabolic side effects.
Although it has taken nearly 30 years to establish what shunting operations do and do not accomplish, several recent reports have shed light on this subject.3–6 All shunting operations are unphysiologic because they deprive the liver of variable quantities of portal blood flow and reroute ammonia-rich colonic blood into the systemic circulation.How do we evaluate published reports on shunts?
There are limitations and criticisms of all retrospective and prospective studies on shunting operations. These limitations include study design, patient selection, and manner and type of follow-up. Results of most retrospective studies have shown that shunts were inherently “protective,” and that the absence of randomization and control patients is a serious drawback. The limitation of most of the randomized studies is that the patients are of the lower socioeconomic groups and,. . .