Following endoscopic therapy of variceal bleeding, the issue of when to refeed these patients has rarely been investigated. This may imply that feeding is generally regarded as a negligible factor in the management of bleeding varices. On the other hand, it is usually believed that systematic fasting is required in case of patients with upper gastrointestinal bleeding.

Some researchers in their studies demonstrated that immediate refeeding has similar outcomes to delayed refeeding among patients with low risk of nonvariceal bleeding. The situation of variceal bleeding is quite different from that of peptic ulcer bleeding as demonstrated by studies of other researchers who made a review on feeding of patients with upper gastrointestinal bleeding, and recommended that feeding should be delayed for at least 48 hours after endoscopic therapy because early refeeding may cause a shift in blood flow to the splanchnic circulation, which in turn could lead to an increase in pressure and an increased risk of rebleeding from the varices.

The other important reason of delay in feeding may be ascribed to the fear of occurrence of early rebleeding induced by refeeding. In addition, repeated endoscopic examination and therapy may be required in patients with very early rebleeding. The decision to delay feeding is usually based on clinicians 'experience or experts' opinion rather than evidence based. That is why we planned this study.

There has not been much studies conducted to address this issue however most of the recommendations are made in view of preventing any complications following procedures. Stiegmann and Goff 4 were the first to employ EVL to treat esophageal varices. They did not mention any specific restriction of feeding on patients receiving band ligation however they did mentioned that following the band ligation of varices, the ligated varices and tissues may evoke transient dysphagia to solid food.5 A study conducted by Gin Ho Lo et al 6 demonstrated that early feeding with a liquid diet in conscious patients after successful endoscopic therapy of varices can shorten the hospital stay. The two groups which were created in this study were the early-feeding group and the delayed-feeding group. Patients in the early-feeding group were asked to fast for only 4 hours following endoscopic treatment. Subsequently, a liquid diet (fruit juice, soybean juice, milk, rice in liquid form) was instituted for 3 days. Additionally, <500 cc intravenous fluid with proper electrolyte supplement per day was administered. Thereafter, a soft diet was provided for 3 days, after which a regular diet was resumed since the seventh day after endoscopic treatment. Patients in the delayed-feeding group were asked to absolutely fast for 48 hours after endoscopic treatment, and 1500 cc/day intravenous fluids (normal saline or glucose water) with proper electrolytes were administered for 2 days. After 2 days of fasting, a liquid diet was given for 1 day, and subsequently, a soft diet was given for 3 days, and then a regular diet was instituted on the seventh day after endoscopic treatment. If rebleeding occurred within 7 days of endoscopic therapy, patients in both groups were again asked to fast for 48 hours, and then put on a liquid diet for 1 day followed by a soft diet for 4 days.

Treatment failure was defined as failure to control acute bleeding episodes or very early rebleeding or death within 5 days. Failure to control acute variceal bleeding was defined as the occurrence of any of the following events within 48 hours of enrollment, based on the modified criteria of the Baveno III consensus:7 (1) fresh hematemesis after enrollment; (2) sudden onset of reduction in blood pressure of _20 mmHg and/or an increase in pulse rate of 20 beats/minute with 2 g drop in hemoglobin; (3) transfusion of four units of blood required to increase the hematocrit to above 27% or hemoglobin to above 9 g/dL; and (4) death. Very early rebleeding was defined as when the criteria for failure to control acute variceal bleeding occurred between 48 hours and 120 hours after enrollment in patients achieving control of acute bleeding. Control of acute bleeding (initial hemostasis) was defined as when the criteria for failure did not occur within 48 hours of enrollment. A 5-day hemostasis was defined as when the criteria for failure to control acute variceal bleeding and very early rebleeding did not occur within 5 days of enrollment. A nasogastric tube was not routinely inserted after initial endoscopy.

The rebleeding rates associated with EVL varied greatly between studies, and rates ranging from 2% to 54% have been recorded.8,9 Aside from other reasons, this discrepancy is very likely to be related to the variation in timing of refeeding after EVL. However, this factor was rarely mentioned in the methodology of these studies A review article by Xavier He´buterne 10 concluded that In patients hospitalized for acute upper gastrointestinal bleeding due to an ulcer with high risk of rebleeding (Forrest I-IIb) or with variceal bleeding it is recommended to wait at least 48 h after endoscopic therapy before initiating oral or enteral feeding. In case of ulcer with low risk of rebleeding (Forrest IIc and III) or in patients with gastritis, Mallory-Weiss, oesophagitis, or angiodysplasia, there is no need to delay refeeding, and they can be fed as soon as tolerated. Understanding the cause of the diagnosis is always necessary to adapt nutrition in patients with upper gastrointestinal bleeding.

Inclusion Criteria

1. Presence of Portal Hypertension with or without cirrhosis.
2. 18 and 80 years of age.
3. Prophylactic EVL
4. EVL was performed as maintenance treatment for Past history of bleeding varices
5. Emergency endoscopic EVL for acutely bleeding varices.

Exclusion Criteria:

1. Septicemia
2. Patients with advanced hepatocellular carcinoma (HCC)
3. Failure to control variceal bleeding by emergency endoscopic therapy.
4. Uncooperative patient
5. Patient with endotracheal intubation.
6. Ever received EIS or EVL within 1 month prior to index bleeding.
7. Child Pugh's scores >13.
8. Presence of Overt hepatic encephalopathy

Based on earlier study done by Gin-Ho Lo et al where the prevalence of UGI bleed in early refeeding group was 33 % and late refeeding group was 10 % a sample size required to detect any significant difference was found to be 88. Adding to which 10 % of attrition a total sample size of 100 was taken.