Protocol for the Evaluation of an Outcomes Based Care for Vulnerable Older People

Introduction: The North East Wyong region consists primarily of an aged population with the lowest socioeconomic status on the Central Coast in Australia. There is a high rate of hospital admissions in aged patients with chronic diseases, and this fragmented, complex and costly healthcare behooves alternative care to better manage this patient group in a community or sub-acute setting. In 2017, Central Coast will implement an outcomes based care program. It represents the first payment structure based upon the delivery of outcomes rather than activity for integrated care in Australia. Aim: We have developed a research protocol to evaluate the process and cost-effectiveness of implementing the outcomes based payment model, and to explore the potential unintended consequences from non-government provision of coordinated care. Target population and stakeholders: The overall objective is to provide coordinated access to cost effective care to vulnerable patients aged 65 years and above with two or more chronic conditions who are at risk of unplanned hospitalisation. The programme will commence in January 2017 for one year. The design is a non-randomised intervention study. The intervention and control group of patients were identified as being ‘at risk’ of hospital admission through a risk stratification algorithm used by Central Coast Local Health District (CCLHD). The algorithm is based on age, disease and previous hospital admissions. Two private providers will provide social care services to the population. The intervention group will be attached to two general practices per provider. A control group of matched patients will allow comparisons between the integrated care programme and usual care. Providers are incentivised to improve patient health with payment managed through a shared risk arrangement with CCLHD. Part of the reimbursement to providers will be calculated based on the total number of saved unplanned public hospital inpatient days. Methods: The overall design is a mixed-methods study using primarily quantitative and additional qualitative data from general practices and service providers, patient reported outcomes, patient and carer interviews, and interviews and focus group data from staff in general practices and service providers. A process evaluation will examine whether the program was implemented as planned, stakeholder response to the programme, characteristics of enrolled and drop-out patients, accuracy of the risk stratification algorithm in predicting unplanned public hospital inpatient days, impact of the program and any changes in service delivery during the study. An economic evaluation to be undertaken from a societal perspective, will provide the financial cost of the outcomes based care model according to utilisation of services and cost per patient according to clinical condition, and the rate of avoidable hospitalisations. Costs to the patient, CCLHD, federal government and the private sector (providers, NGOs and GPs) are included, where the data allows. Transferability: This evaluation will generate a better understanding of the factors that can influence the implementation, feasibility and sustainability of adapting outcomes based care to fit other healthcare and social contexts within Australia and internationally. It will also highlight the usefulness of a purpose built algorithm to estimate unplanned public hospital inpatient bed days.