PTU-133 Predicting Mortality in Alcholic Hepatitis; a Comparison in Different Scoring Systems

    loading  Checking for direct PDF access through Ovid

Abstract

Introduction

Alcoholic hepatitis is a severe presentation of alcoholic liver disease. In its most severe form (with a Maddray score >32) the 1 month mortality is 35%. There are currently a few methods used to decide if continuation of corticosteroid therapy is beneficial, the Lille’s Score and the presence of a 25% reduction in the serum bilirubin between day 1 and day 7. We assess the effectiveness of these scoring systems in assessing 6 month mortality.

Methods

All patients admitted with a diagnosis of alcoholic hepatitis with a Maddray score > 32 with no evidence of sepsis over a 2 year period (2010–2012) were identified. All the notes were analysed and data collected by a F1 using a standardised profroma. Patients were placed into 4 groups: Group 1= <25% reduction in bilirubin, low Lille score, Group 2= <25% in bilirubin, high Lille score, Group 3= >25% reduction in bilirubin, low Lille score and Group 4= >25% reduction in bilirubin, high Lille score.

Results

Overall 77 patients were included, at 6 months 21 died (27%). The mean age was 48 yrs (Range 27–67 years). At day 1 there was biochemical parameters consistent with significant liver disease (Madray Score: 68 (Range 34–169), Albumin 29.9 ± 4.9, INR: 2.0 ± 0.5, Bilirubin: 235 ± 135). 77 (100%) patients received nutritional support, Vitamin B and Thiamine. The baseline INR (Alive: 1.85 ± 0.47, Died: 2.31 ± 0.59 p = 0.001) and Albumin (Alive: 30.6 ± 4.6, Died: 28.1 ± 5.5 p = 0.04) were significantly deranged in patients who died at 6 months. There was no significant difference in the baseline Urea (Alive: 3.81 ± 3.54, Died: 4.26 ± 3.46 p = 0.610), Creatinine (Alive: 64.3 ± 45.9, Died: 79.5 ± 59.6 p = 0.232) and Bilirubin (Alive: 237.8 ± 144.6, Died: 228.1 ± 110.3 p = 0.779) in patients who died at 6 months.

Results

There were no patient that fell into group 4. In the other 3 groups there were similar numbers of patients (Group 1: 29 patients, 6 month mortality 17%, Group 2: 23 patients, 6 month mortality 57%, Group 3: 25 patients, 6 month mortality 12%). Kaplan Meier survival curves were created for these 3 groups and is shown in Figure 1 below.

Conclusion

In this study factors suggesting poor liver synthetic function (INR and Albumin) were associated with 6 month mortality. There was a significantly worse outcome with a high Lille score compared to a low Lille score. There was very little effect of a greater than 25% reduction in bilirubin on mortality at 6 months. From this study we would suggest that the Lille score is used to accurately predict a poor outcome.

Disclosure of Interest

None Declared.

Related Topics

    loading  Loading Related Articles