Development and Management of Networks of Care at the End of Life (the REDCUIDA Intervention): Protocol for a Nonrandomized Controlled Trial

Background: End-of-life needs can be only partly met by formalized health and palliative care resources. This creates the opportunity for the social support network of family and community to play a crucial role in this stage of life. Compassionate communities can be the missing piece to a complete care model at the end of life. Objective: The main objective of this study is to evaluate the REDCUIDA (Redes de Cuidados or Network of Care) intervention for the development and management of networks of care around people with advanced disease or at the end of life. Methods: The study is a 2-year nonrandomized controlled trial using 2 parallel groups. For the intervention group, we will combine palliative care treatment with a community promoter intervention, compared with a control group without intervention. Participants will be patients under a community palliative care team’s supervision with and without intervention. The community promotor will deliver the intervention in 7 sessions at 2 levels: the patient and family level will identify unmet needs, and the community level will activate resources to develop social networks to satisfy patient and family needs. A sample size of 320 patients per group per 100,000 inhabitants will offer adequate information and will give the study 80% power to detect a 20% increase in unmet needs, decrease families’ burden, improve families’ satisfaction, and decrease the use of health system resources, the primary end point. Results will be based on patients’ baseline and final analysis (after 7 weeks of the intervention). We will carry out descriptive analyses of variables related to patients’ needs and of people involved in the social network. We will analyze pre- and postintervention data for each group, including measures of central tendency, confidence intervals for the 95% average, contingency tables, and a linear regression. For continuous variables, we will use Student t test to compare independent samples with normal distrib