ABSTRACT
OBJECTIVES: To comprehensively assess the effect of a living will on end-of-life care. DESIGN: Systematic review with narrative analysis following Preferred Reporting Items for Systematic Reviews and Meta-Analyses methodology. PARTICIPANTS: All interventional and observational studies were included, excepting those with fictive cases. Included studies were conducted in adults with and without living wills, excluding individuals with specific psychiatric living wills. MEASUREMENTS: Two authors independently extracted study and participant characteristics and outcomes related to end-of-life care (place of death, hospitalization and intensive care unit management, life-sustaining treatments, restricted care). Risk of bias was assessed using the Risk Of Bias In Non-randomized Studies of Interventions tool. RESULTS: From 7,596 records identified, 28 observational studies were included, 19 conducted in the United States, 7 in Europe, and 1 each in Canada and Australia. Place of death was assessed in 14 studies, life-sustaining treatments in 13, decision for restricted care in 12, and hospitalization in 8. Risk of bias was serious for 26 studies and moderate for 2. Twenty-one studies reported significantly less medical management for individuals with a living will, 3 reported more medical management, and the difference was not significant in 37. CONCLUSION: Methodological quality of included studies was insufficient to offer reliable results. The effect of living wills appears limited in view of the importance and direction of potential biases. Further studies including larger populations, considering main confounding factors, and documenting the real presence of a living will in medical records are necessary to reach stronger conclusions on the effect of living wills on end-of-life care. J Am Geriatr Soc 67:164-171, 2019.