End-of-Life Care Preferences of Older Patients with Multimorbidity: A Mixed Methods Systematic Review

Abstract

Unpredictable disease trajectories make early clarification of end-of-life (EoL) care preferences in older patients with multimorbidity advisable. This mixed methods systematic review synthesizes studies and assesses such preferences. Two independent reviewers screened title/abstracts/full texts in seven databases, extracted data and used the Mixed Methods Appraisal Tool to assess risk of bias (RoB). We synthesized findings from 22 studies (3243 patients) narratively and, where possible, quantitatively. Nineteen studies assessed willingness to receive life-sustaining treatments (LSTs), six, the preferred place of care, and eight, preferences regarding shared decision-making processes. When unspecified, 21% of patients in four studies preferred any LST option. In three studies, fewer patients chose LST when faced with death and deteriorating health, and more when treatment promised life extension. In 13 studies, 67% and 48% of patients respectively were willing to receive cardiopulmonary resuscitation and mechanical ventilation, but willingness decreased with deteriorating health. Further, 52% of patients from three studies wished to die at home. Seven studies showed that unless incapacitated, most patients prefer to decide on their EoL care themselves. High non-response rates meant RoB was high in most studies. Knowledge of EoL care preferences of older patients with multimorbidity increases the chance such care will be provided.

Keywords: elderly; end of life care; multimorbidity; patient centered care; patient preferences.

Figure 1

Forest plot comparing the willingness to receive life-sustaining treatments: Shows results (%) from studies where the context was a hypothetical scenario, a real scenario or regardless of the scenario patients were presented with when end-of-life care preferences were assessed.

Figure 2

Willingness to receive life-sustaining treatments in different scenarios: (a) Compares preferences (%) from two studies assessing the willingness to receive LSTs depending on the risk of death, cognitive or functional impairment (%); (b) Compares preferences (%) from three studies assessing the willingness to receive LSTs depending on life extension and treatment burden.

Figure 3

Forest plots comparing willingness to receive cardiopulmonary resuscitation: (a) Shows results (%) from studies where the context was a real scenario or both scenarios and includes the results from Modes et al. assessing patients’ preferences in their current state of health; (b) Shows results (%) from studies where the context was a real scenario or both scenarios and includes the results from Modes et al. assessing patients’ preference in a hypothetical state of dependency.

Figure 4

Forest plots comparing willingness to receive mechanical ventilation: Shows results (%) from studies where the context was a real scenario, a hypothetical scenario or both scenarios.

Figure 5

Forest plot comparing the preference for home as a place of end-of-life care: Shows results (%) from studies where the context was a real scenario or both scenarios.

Figure 6

Preferences regarding place of end-of-life care: Shows results (%) from studies that assessed preferences regarding the place of end-of-life care.Two studies [38,45] assessed patients’ preferences for ICUs as a place to receive EoL care (Figure 6). Thirty-six percent of patients with advanced cancer and a life expectancy of less than six months were against dying in an ICU [45], compared with 91% of older adults with multimorbidity but a life expectancy over six months, who considered dying in an ICU to be a ‘bad death’ [38].