Abstract
Background and Aims: Liver transplantation is the primary treatment choice for alcohol liver disease. However, alcohol relapse is a particular problem for this population with a reported 10–50% risk. This review aims to identify alcohol relapse variables, and establish the effectiveness and explore the active ingredients of psychosocial interventions in preventing alcohol relapse, in alcohol liver disease patients pre or post-liver transplant. Methods: Mixed method systematic review involving three parallel syntheses: (1) alcohol relapse variables; (2) psychosocial intervention effectiveness; (3) active ingredients of these psychosocial interventions (i.e. a component analysis). Medline, CINAHL, EMBASE, and PsychInfo were searched in November 2013 for published literature. Grey literature searched in Web of Science, Clinical Trials Register, and Electronic Theses Online Service. Synthesis 1: Systematic search for and appraisal of prospective studies, retrospective studies, and cross-sectional surveys. Syntheses 2/3: Systematic search for and appraisal of ‘randomised controlled trials’, ‘controlled before and after studies’, and ‘before and after studies in a single group’. Results: 23 papers included: 10 cohort, 11 case–control, 1 qualitative, and 1 randomised controlled trial. Five variables out of nineteen were alcohol relapse predictors (i.e. Synthesis 1): (1) <12 months pre-transplant abstinence; (2) presence of children; (3) poor pre-transplant psychosomatic evaluation; (4) non-compliant with post-transplant treatment plan; (5) active insurance policies at transplant. One psychosocial intervention paper did not report treatment effectiveness (i.e. Synthesis 2). The remaining three papers reported relapse rate reduction: Alcohol Addiction Unit (odds ratio 0.23), Structured Management (odds ratio 0.32), pre- and post-transplant Substance Abuse Treatment (odds ratio 0.27 compared to no substance abuse treatment; odds ratio 0.23 compared to pretransplant substance abuse treatment only). With confidence intervals not reported, the uncertainty level around the odds ratio is unclear. Furthermore, a theoretical basis was not discussed; thus, the active ingredients could not be identified (i.e. Synthesis 3). Conclusions: Randomised controlled trials to further investigate the predictive validity of the five main variables and ascertain the long-term benefits of the tentative yet promising results of the most effective intervention i.e. structured management.