It has been observed that the portal pressure in cirrhotic patients shows circadian variation and is at a maximum at night. Our aims in this study were a) to evaluate the frequency of upper gastrointestinal (UGI) bleeding from esophageal/gastric varices or portal congestive gastropathy during the day, evening or night shift, and b) to correlate the frequency of bleeding during various shifts with severity of liver disease.At our institution, during the period from January 1997 to June 1995, 15,000 gastrointestinal (GI) endoscopic reports and consultations were evaluated. Episodes of bleeding due to portal hypertension (PH), i.e., esophageal varices, gastric varices or congestive gastropathy, were noted. The time of occurrence of bleeding from PH in various shifts, e.g., day shift (DS), evening shift (ES) or night shift (NS) was determined. The severity of liver disease was classified as A, B or C according to Child's classification.There were 221 episodes of UGI bleeding due to PH in 144 patients. 75.5% of patients presented with hematemesis, whereas 24.5% presented with melena (p < 0.025). Bleeding started during the DS in 39/221 (17.6%), during the ES in 122/221 (55.2%), and during the NS in 60/221 (27.1%). There were 29 (13.1%) patients in Child's A, 75 (33.9%) in Child's B, and 117 (52.9%) in Child's C. Among the 122 bleeders during ES, 85 (69.6%) were in category C. It seems that in these patients UGI bleeding is more frequent during the ES compared to DS and NS (p < 0.0001; chi square; 95% CI: 0.52-0.58). Child's C patients are more likely to bleed during ES compared to Child's A or B patients (p < 0.0001; Odds Ratio: 4.8%; 95% CI: 2.7-8.5).1) The majority of the cirrhotic patients with PH who develop UGI bleeding present with hematemesis rather than melena. 2) The bleeding in these patients is more likely to occur during the ES. 3) Child's C patients are more likely to bleed during the ES.
The usual method of estimating probability of infectivity for donors implicated in more than one case of post‐transfusion hepatitis fails to take adequate account of the donor's past history and underestimates the risk of infectivity. A corrected method of calculation is presented and shown to give substantially different results.
