End-of-Life Decisions in Individuals Dying with Dementia in Belgium

ObjectivesTo describe the prevalence and characteristics of end-of-life decisions in individuals dying with dementia in Belgium. DesignRetrospective mail survey. SettingBrussels and Flanders, Belgium. ParticipantsPhysicians certifying a large random sample of deaths (n=8,627). MeasurementsEnd-of-life decision-making in individuals dying with dementia (n=361) was compared through bivariate and multivariable analyses with decision-making in individuals dying from cancer (n=1,276) as reference group. ResultsResponse rate was 54.7%. The intense pain and symptoms of individuals with dementia was alleviated less often than those of individuals with cancer (odds ratio (OR)=0.50, P<.001), and individuals with dementia were more likely to have life-prolonging treatment withheld or withdrawn (OR=1.40, P=.048). Five individuals with dementia (1.3%) requested euthanasia but did not receive it; 60 (4.7%) individuals with cancer who requested euthanasia received it, and 48 (3.8%) who requested it did not (OR=0.04, P=.047). Individuals with dementia rarely had capacity to participate in decision-making, which more often involved their families than in individuals with cancer (OR=1.99, P=.009). Little prospect of improvement (59.7%), pointless life prolongation (54.5%), poor quality of life (46.2%), and prevention of further suffering (37.6%) were important reasons for end-of-life decisions in individuals with dementia. ConclusionEnd-of-life decision-making in individuals dying with dementia differs from that in individuals with cancer, more often involving forgoing of life-prolonging treatment and less often involving intensifying pain and symptom treatment. Considerations typically involve the prospects of the individual with dementia as opposed to the severity of their present situation. Optimal processes of advance care planning may address the burdensome decision falling to physicians and family regarding when to allow the individual to die.