Résumé
Injection sclerotherapy has become an effective modality for treatment of bleeding esophageal varices inspite of advancement in equipment, sclerosant solutions, and techniques. Injection sclerotherapy induced esophageal strictures remain a significant cause of patient morbidity. To evaluate the risk factors and prognosis, the records of 80 patients who underwent injection sclerotherapy over a 3 year period were reviewed. In all cases a free hand injection technique, flexible endoscopes and sodium morrhuate were used. Multiple risk factors were studied including patient age, Child's classification, previous bleeding attacks, aetiology of varices, amount of sclerosant used, and number of injection. During a mean follow up period 200 days, 25 patients [50 percent] died and 8 patients [16 percent] developed symptomatic strictures. The cumulative amount of sclerosant used [9 +/- 80m]] and the number of injection sclerotherapy [1 +/- 7] required in the stricture group was significantly greater than with the nonstricture group [2 +/- 45 and 4.0 +/- 2] respectively. The risk of stricture formation did not correlate to volume of sclerosant injected per treatment, cause of varices, number of previous bleeds or Child's class. Dilations was recurred for treatment of established stricture. Six patients [75 percent] required repeated dilations for four times or less. It is concluded that the risk of injection sclerotherapy can be predicted from the number of injection sclerotherapy and the cumulative volume of sclerosant required to obliterate the varices, management of strictures usually requires repeated dilations. The high mortality rate in these patients appears to reflect the severity of liver disease as 12/16 [75 percent] of deaths occurred in Child's C patients