Approche non Endoscopique du Diagnostic des Varices Œsophagiennes d’Origine Cirrhotique dans une Population d’Afrique Noire Subsaharienne

2015 
RESUME OBJECTIFS. Identifier des parametres cliniques et para cliniques pour l’approche non endoscopique du diagnostic des varices œsophagiennes chez les patients atteints de cirrhose, en contexte sub-saharien. MATERIELS ET METHODES. Les dossiers de patients cirrhotiques camerounais examines au Centre Hospitalier et Universitaire de Yaounde entre mars 2013 et novembre 2014 ont ete revus. Les donnees cliniques, biologiques, echographiques et ont ete analysees et comparees a la severite des varices œsophagiennes retrouvees par endoscopie. La methode de classification et regression par carte a permis de construire un arbre de prediction des varices œsophagiennes. RESULTATS. Nous avons inclus 98 patients cirrhotiques d’âge compris entre 14 et 80 ans. Les hepatites virales B et C ont ete retrouvees chez 78 patients (79,59%). 53 patients (54,08%) etaient classes Child-Pugh C. La prevalence des varices œsophagiennes etait de 92% (90/98) avec 63,27% de grade III. Les variables independamment associes a la severite des varices œsophagiennes etaient les taux de prothrombine, d’hemoglobine, d’albumine, de leucocytes, de creatinine serique et la taille de la rate. Ces facteurs ont ete utilises pour construire un arbre de prediction de survenue des VO. CONCLUSION. A l’aide d’un arbre decisionnel integrant la taille de la rate et cinq facteurs biologiques, il est possible de predire la severite des varices œsophagiennes post-cirrhotiques en contexte sub-saharien. SUMMARY OBJECTIVES. To identify some clinical and para clinical parameters for non-endoscopic screening of esophageal varices in patients with liver cirrhosis, in sub-Saharan setting. MATERIALS AND METHODS. The medical records of black African Cameroonian cirrhotic patients received between March 2013 and November 2014 in the University Teaching Hospital of Yaounde were reviewed. Clinical, biological and sonographic findings were analyzed and compared to the severity of endoscopically diagnosed esophageal varices. The classification and regression tree method was used to build a prediction tree of esophageal varices. RESULTS. We studied 98 cirrhotic patients aged 14-80 years. Viral hepatitis B and C were found in 78 (79.59%) patients. Cirrhosis was graded Child-Pugh C in 53 patients (54.08%). The prevalence of esophageal varices was 92% (90/98) with 63.27% of grade III. The factors independently associated with the severity of esophageal varices were prothrombin time, hemoglobin level, albumin, leukocytes, serum creatinin level and the size of the spleen. We used these factors to build a predictive tree for esophageal varices. CONCLUSION. A prediction tree including the spleen size and five biological factors could help to predict the severity of post-cirrhotic esophageal varices in sub-Saharan setting without systematic use of endoscopy.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    19
    References
    1
    Citations
    NaN
    KQI
    []