There is a variation in dialysis withdrawal rates, but reasons for this variation across European countries are largely unknown. We therefore surveyed nephrologists' perceptions of factors concerning dialysis withdrawal and palliative care and explored relationships between these perceptions and reports of whether withdrawal actually occurred in practice.Methods
We developed a 33-item electronic survey, disseminated via an email blast to all European Renal Association–European Dialysis and Transplant Association (ERA-EDTA) members. In our data analyses, we distinguished those respondents who reported occurrence from those reporting no dialysis withdrawal in their unit. With multilevel logistic regression, we investigated the association between respondents' characteristics and perceptions and whether they reported occurrence of dialysis withdrawal or not.Results
Five hundred and twenty-eight nephrologists from 45 countries completed the questionnaire; 42% reported occurrence of withdrawal in their unit in the past year, and 56% perceived that stopping life-prolonging treatment in terminally ill patients was allowed. Few respondents reported presence in their unit of protocols on withdrawal decision making (7%) or palliative care (10%) or the common involvement of a geriatrician in withdrawal decisions (10%). The majority stated that palliative care had not been part of their core curriculum (74%) and that they had not recently attended continuous medical education sessions on this topic (73%). Respondents from Eastern and Southern Europe had a 42 and 40% lower probability, respectively, of reporting withdrawal compared with those from North European countries. Working in a public centre [odds ratio (OR), 2.41; 95% confidence interval (CI), 1.36–4.25] and respondents' perception that stopping life-prolonging treatment in terminally ill patients was allowed (OR, 1.96; 95% CI, 1.23–3.12), that withdrawal decisions were commonly shared between doctor and patient (OR, 1.97; 95% CI, 1.26–3.08) and that palliative care was reimbursed (OR, 1.81; 95% CI, 1.16–2.83) increased the odds of reporting occurrence of withdrawal.Conclusion
Reports of dialysis withdrawal occurrence varied between European countries. Occurrence reports were more likely if respondents worked in a public centre, if stopping life-prolonging treatments was perceived as allowed, if withdrawal decisions were considered shared between doctors and patients and if reimbursement of palliative care was believed to be in place. There is room for improvement regarding protocols on withdrawal and palliative care processes and regarding nephrologists' training and education on end-of-life care.