Factors related to 30th day mortality in upper digestive tract hemorrhage in cirrhotic patients
Poynard, T.; Chaput, J.C.; Mary, J.Y.; Scolaro, M.; Buffet, C.; Etienne, J.P.
Gastroenterologie Clinique et Biologique 4(10): 655-665
1980
ISSN/ISBN: 0399-8320 PMID: 6969199 Document Number: 163532
Uni- and multidimensional analyses were used to calculate the short-term prognostic value of 17 clinical and laboratory variables estimated during 116 episodes of digestive hemorrhage occurring in patients with cirrhosis. Unidimensional analysis demonstrated that the subjects who died within the 1st 30 days (n = 40), as compared with those who did not (n = 76), had a significantly higher incidence of ascites (P < 0.01), higher serum bilirubin (P < 0.01) and more abnormal prothrombin time (P < 0.01). Rupture of esophageal varices as the cause of hemorrhage (P < 0.05) and a lower incidence of aspirin ingestion (P < 0.05) were significantly related to a fatal outcome. Multidimensional analysis demonstrated that the only variable to have an independent unfavorable prognostic value (P < 0.02) was ascites. By multiple regression analysis, age was also associated with a poor prognosis (P < 0.05), although this relationship had remained concealed with unidimensional analysis. The low prognostic values of those factors reported in the published literature are discussed. The prognosis of digestive hemorrhage in cirrhosis is evidently multifactorial. The classifications published until now, apparently do not adequately take into account the clinically available information. Only 2 clinical variables, ascites and age, evidently had significant independent explanatory values.