Methods. We performed a survey in a sample of 1745 physicians of the Paris region including gastroenterologists, transplanters, general practitioners and occupational physicians. Using a standardized mailed questionnaire describing an hypothetical scenario, physicians were asked to allocate 100 liver transplants between two competing groups of patients equivalent for every characteristic except for cirrhosis origin (i.e. alcohol-related or autoimmune). The number of liver transplants allocated to each patient group was classified in <50 and ≥ 50. Responses to 9 questions were combined into a composite score of personal responsibility for illness. Factors associated with the allocation of <50 liver transplants to patients with alcohol-related cirrhosis were analyzed using regression logistic models.

Results. Among the 537 respondents (response rate: 30.9 %), 63.0 % were male, 45.3 % aged between 45 and 54, 46.9 % general practitioners and 53.3 % reported at least one substance consumption ; 55.3 % allocated <50 liver transplants to patients with alcohol-related cirrhosis. This was independently associated with physician opinion that patients were personally responsible for their illness (odds ratio [OR] = 4.2, 95 % confidence interval [CI] = 1.8-10.0, increasing to OR=126.3, 95 % CI = 34.0-468.3 for the highest composite score), being a general practitioner (OR = 3.1, 95 % CI = 1.8-5.6) and scenario misinterpretation (OR = 1.8, 95 %CI = 1.1-3.0).

Conclusions. More than half physicians would have limited access to liver transplant of patients with alcohol-related cirrhosis. This occurred especially in general practitioners, physicians having misunderstood the scenario or thinking that patients with excessive alcohol consumption are responsible for their illness. However, education of and information to physicians, especially general practitioners, may improve the access to liver transplantation of patients that could benefit from this procedure