INTRODUCTION: The variceal bleeding, main complication of portal hypertension during cirrhosis, is associated with high early mortality riskestimated between 15 and 20%. This highlights the necessity of predictive models that allow identifying high-risk patients raising the issue of amore aggressive therapeutic care. OBJECTIVE: To assess the performance of four scores for the prediction of cirrhotic patients' high early mortality risk due to digestive hemorrhageand to compare them to the Child-Pugh reference score. METHODS: We collected 87 cirrhotic patients admitted to the Gastroenterology Department of Charles Nicolle Hospital for a high digestivehemorrhage by rupture of gastric or esophageal varicose veins. RESULTS: 56 men and 31 women were included in this study. The average value of Rockall, Glasgow Blatchford, MELD and MELD-Na scores,was respectively equal to 6.19, 10.91, and 17.6 and at 20. Early mortality was 30%. The average value of all the scores was significantly higherwith the prematurely deceased patients (p<0.001). The MELD-Na score had higher sensitivity and specificity for the prediction of prematuremortality compared to the other scores but without statistical significantly difference (Area under the ROC curve: MELD-Na=0.867, p<0.001;Child-Pugh=0.809, p<0.001; Rockall=0.777, p=0.001; Glasgow-Blatchford=0.761, p<0.001; MELD=0.838, p<0.001). The predictive value of thecut-off MELD-Na score was equal to 19 with a sensitivity of 70% and a specificity of 82%. CONCLUSION: The studied four scores had a good predictive value of early mortality risk by varicose digestive hemorrhage with cirrhotic patients.
INTRODUCTION: The variceal bleeding, main complication of portal hypertension during cirrhosis, is associated with high early mortality riskestimated between 15 and 20%. This highlights the necessity of predictive models that allow identifying high-risk patients raising the issue of amore aggressive therapeutic care. OBJECTIVE: To assess the performance of four scores for the prediction of cirrhotic patients' high early mortality risk due to digestive hemorrhageand to compare them to the Child-Pugh reference score. METHODS: We collected 87 cirrhotic patients admitted to the Gastroenterology Department of Charles Nicolle Hospital for a high digestivehemorrhage by rupture of gastric or esophageal varicose veins. RESULTS: 56 men and 31 women were included in this study. The average value of Rockall, Glasgow Blatchford, MELD and MELD-Na scores,was respectively equal to 6.19, 10.91, and 17.6 and at 20. Early mortality was 30%. The average value of all the scores was significantly higherwith the prematurely deceased patients (p<0.001). The MELD-Na score had higher sensitivity and specificity for the prediction of prematuremortality compared to the other scores but without statistical significantly difference (Area under the ROC curve: MELD-Na=0.867, p<0.001;Child-Pugh=0.809, p<0.001; Rockall=0.777, p=0.001; Glasgow-Blatchford=0.761, p<0.001; MELD=0.838, p<0.001). The predictive value of thecut-off MELD-Na score was equal to 19 with a sensitivity of 70% and a specificity of 82%. CONCLUSION: The studied four scores had a good predictive value of early mortality risk by varicose digestive hemorrhage with cirrhotic patients.
Authors: Aurélia Rakotondrainibe; Nadia M P Rahanitriniaina; Harifetra M R Randriamizao; Jasper G Raelison; Rado M Ramanampamonjy; Andriambelo T Rajaonera; François Sztark Journal: Afr J Emerg Med Date: 2020-07-21