BackgroundWorldwide, the number of individuals lacking the mental capacity to participate in decisions about their own healthcare is increasing. Due to the ageing global population and advancing medical treatments, there are now many more people living longer with neurological disorders, such as dementia, acquired brain injuries, and intellectual disabilities. Many of these individuals have feeding difficulties and may require artificial nutrition. However, little is known about the decision-making process; the evidence base is uncertain and often ethically complex. Using the exemplar of artificial nutrition, the objective of this review is to examine how treatment decisions are made when patients are at risk of lacking capacity.Methods and FindingsWe undertook a systematic review according to PRISMA guidelines to determine who was involved in decisions, and what factors were considered. We searched PubMed, AMED, CINAHL, EMBASE, PsychINFO, and OpenSigle for quantitative and qualitative studies (1990–2011). Citation, reference, hand searches and expert consultation were also undertaken. Data extraction and quality assessment were undertaken independently and in duplicate. We utilised Thomas and Harden’s ‘Thematic Synthesis’ for analysis. Sixty-six studies met inclusion criteria, comprising data from 40 countries and 34,649 patients, carers and clinicians. Six themes emerged: clinical indications were similar across countries but were insufficient alone for determining outcomes; quality of life was the main decision-making factor but its meaning varied; prolonging life was the second most cited factor; patient’s wishes were influential but not determinative; families had some influence but were infrequently involved in final recommendations; clinicians often felt conflicted about their roles.ConclusionsWhen individuals lack mental capacity, decisions must be made on their behalf. Dynamic interactive factors, such as protecting right to life, not unnecessarily prolonging suffering, and individual preferences, need to be addressed and balanced. These findings provide an outline to aid clinical practice and develop decision-making guidelines.