Evaluating an integrated care programme for frail elderly using Multi-Criteria Decision Analysis. The design of a case study in SELFIE

Abstract

Background : There is a rising need for integrated care for frail elderly to support independent living. Although many initiatives to integrate care arise, decisions to reimburse such care remain difficult. Decision-makers require evidence on the effectiveness. However, there are many methodological challenges to overcome. One of them is the choice of outcome measures. Many previous studies were not able to detect improvements, potentially because they focus on traditional health-related outcomes instead of what matters most to the elderly. Furthermore, evaluations generally include numerous outcomes, which may show conflicting results. This makes it difficult for policymakers to compare and make trade-offs . Aim : In the Horizon2020 project SELFIE we aim to evaluate promising integrated care programmes using Multi-Criteria Decision Analyses MCDA that both measure and weigh a broad array of outcomes. One of these programmes is the Care Chain Frail Elderly CCFE in Southeast Brabant NL, an intensive integrated care programme for the frailest elderly living at home. The CCFE aims to support frail elderly by delivering individualised, integrated care, so that elderly can live at home independently. The aim of this abstract is to discuss the study design and the choices in creating that design. Methods : A prospective quasi-experimental evaluation study was designed to compare the CCFE with usual care expected N=200/group. Outcomes are measured at baseline, and 6 and 12 months thereafter. Selection of outcomes was guided by focus groups with persons with multi-morbidity. Outcomes were classified according to the Triple Aim. Simultaneously, we set up a study to determine weights reflecting the importance of these outcomes from the perspective of different stakeholders: patients, partners, professionals, payers, and policymakers n=150 per group. Two different methods, i.e. Discrete Choice Experiment and Swing Weighting were used. Results : Patients in the control group were recruited from different GPs than patients in the intervention group, but they operate in the same region. To ensure similarity of patients in both groups, GPs in the control group adopted the same method to identify frail elderly as in the intervention group, but they don’t provide the integrated care programme. A ‘core set’ of eight relevant outcomes to evaluate integrated care is specified: physical functioning, psychological well-being, social relationships and participation, enjoyment of life, resilience, person-centeredness, continuity of care, and total health- and social care costs. This is supplemented with five outcomes specifically relevant for frail elderly: autonomy, burden of medication, informal caregiver burden, falls, and long-term institution admissions. To minimise the risk of excluding the frailest elderly that are not able to complete questionnaires, data collection is being done by trained interviewers during home visits. Conclusion : We present an innovative quasi-experimental study design to evaluate integrated care for frail elderly using MCDA. Among the broad range of innovative outcomes that are particularly relevant for frail elderly were included, such as enjoyment of life and autonomy. Importance-weights for these outcomes were measured from five different perspectives. The MCDA will combine the outcomes with these weights into an overall recommendation on the effectiveness of the frail elderly programme.