Introduction In the UK multicentre, retrospective, real world study, IMPRESS, rifaximin-α (RFX) use in patients with hepatic encephalopathy (HE) significantly reduced hospitalisations and length of stay in the 6 and 12 months post-RFX initiation compared to the respective periods pre-RFX initiation. This post-hoc analysis of the IMPRESS data compared hospital resource use pre- and post-RFX initiation in 2 sub-groups of difficult-to-treat HE patients: those with advanced liver disease or ongoing alcohol misuse.

Method Medical records of patients from 11 UK hospitals who were prescribed RFX for HE between July-2008 and May-2014 were retrospectively reviewed; details of demographic and clinical characteristics, and all-cause hospital admissions were collected in the 6 and 12 months pre- and post-RFX initiation. Patients with baseline MELD score ≥15 or not abstinent at the end of the study period were included in this analysis. Statistical significance of the mean change (standard error of the mean, SEM) was calculated using paired t-test or Wilcoxon test.

Results Only patients alive at the end of the 6 and 12 months RFX-treatment periods were included: 114 and 102, respectively. Amongst these, 33/114 (29%, for the 6 months) and 26/102 (25%, for the 12 months) had baseline MELD ≥15; mean age, 63 years; 70% were male; 66% had alcohol-related liver disease; mean MELD 24. The mean (SEM) number of bed days/patient reduced from 25 (6.0) in the 6 months pre- to 15 (5.5) in the 6 months post-RFX initiation, and from 36 (9.5) in the 12 months pre- to 20 (7.7) in the 12 months post-RFX initiation (p value not significant). At 6 months post-RFX initiation, 15/114 (13%) patients were still actively drinking. At RFX initiation, mean age was 56 years; 73% were male, mean MELD was 19. Despite this, the mean (SEM) number of bed days/patient decreased from 36 (7.9) in the 6 months pre- to 15 (5.4) in the 6 months post-RFX initiation (p=0.048), and the mean of hospitalisations/patient fell from 2.8 (0.8) to 1.2 (0.4) (t-test p=0.059; Wilcoxon test p=0.029). Too few patients with continued alcohol misuse were alive at 12 months to evaluate. Two patients reported adverse events, none serious.

Conclusion In UK clinical practice, treatment with RFX for HE for 6 or 12 months suggested trends in reduced hospital length of stay in patients with advanced liver disease and in those with continued alcohol misuse. However, larger studies are needed to strengthen these findings.

Disclosure of Interest M Hudson Conflict with: Norgine, Conflict with: Sponsored Lectures and advisory board member for Norgine, P. Di Maggio Conflict with: Norgine, R. Cipelli Conflict with: Norgine, Conflict with: pH Associates which was commissioned by Norgine to provide support with study design and management, data analysis and scientific editorial services, R. Aspinall Conflict with: Norgine, Conflict with: Sponsored Lectures and advisory board member for Norgine