PROGNOSTIC UTILITY OF GLASGOW ALCOHOLIC HEPATITIS SCORE (GAHS) IN ALCOHOLIC HEPATITIS

Abstract

Noas Tobias Minz, Jagannath Sadangi, Arjun Kuttikrishnan, Teena Thomas

BACKGROUND Severe alcoholic hepatitis is associated with high mortality, 40-50% mortality at one month has been reported. Many scoring systems have been applied to predict survival in alcoholic hepatitis. Patients with high risk need aggressive treatment. So, an accurate scoring system is essential to predict the patients who are at high risk. The aim of this study was to assess the short term (28 day) prognostic utility of Glasgow Alcoholic Hepatitis Score (GAHS) in alcoholic hepatitis by comparing with mDF and MELD scores. MATERIALS AND METHODS 76 patients with alcoholic hepatitis admitted to medical wards in our hospital, meeting the inclusion and exclusion criteria were assessed within 24 hours. Scoring was done on day one with each scoring system, GAHS score taking parameters of age, total WBC count, serum bilirubin, blood urea and Prothrombin Time; mDF score with Prothrombin Time and serum bilirubin; MELD score with serum bilirubin, PT/INR, and creatinine level. Patients were followed up for 28 days with day-to-day in-hospital analysis of status and by telephonic conversation after discharge. Study Design- Cross-sectional analytical study. RESULTS At 28th day, 40 patients survived, and 36 patients died. The median scores for survivors and non-survivors at 28 days were: for GAHS 7 (range 5-10) and 9 (range 7-11), for mDF 58 (range 2.6-214.8) and 85 (range 28.8-510.8), for MELD 21.8 (range 10-38.8) and 31 (range 11-55) respectively. For the prediction of survival at 28th day, the GAHS score >9 has sensitivity 83.3%, specificity 80%, accuracy 81.57%, p <0.001; mDF score >32 has sensitivity 88.88%, specificity 35%, accuracy 60.52%, p<0.044; mDF >77 has sensitivity 55.55%, specificity 60%, accuracy 57.89%, p>0.05; MELD score >11 has sensitivity 100%, specificity 12.5%, accuracy 53.94%, p>0.05; MELD score >27 has sensitivity 77.77%, specificity 70%, accuracy 73.68%, p <0.001. Other independent factors for high mortality were ascites, encephalopathy, high bilirubin level. CONCLUSION GAHS, easy to calculate at bed side, has high sensitivity, specificity and accuracy in short term (28 days) prediction of mortality in alcoholic hepatitis, superior to mDF and MELD scoring systems.

image