Many countries are currently concerned with the planning of networks of Long-Term Care (LTC), which requires considering a multiplicity of policy objectives and anticipating the impact of key uncertainties. Nevertheless, location-allocation literature has not been modelling key health policy objectives, and the use of stochastic planning models entails low practical usability due to prohibitive computational times. This study tackles these issues by proposing an approach that supports the reorganization of LTC networks (in terms of services location, capacity planning and patients allocation) while exploring different health policy objectives and considering uncertainty within a reasonable computational time, leading to the development of a stochastic multi-objective mathematical programming model – the LTCNetPlanner. The LTCNetPlanner builds upon health economics and policy concepts to model the maximization of health and wellbeing together with cost- and equity-related objectives within location-allocation literature. Concerning uncertainty, a scenario-based stochastic approach is developed and alternative scenario reduction methods enabling a faster model resolution are explored within the LTCNetPlanner. Specifically, it is proposed a novel Morphol-KMG method able to reduce the number of scenarios while accounting for experts’ knowledge. A case study in the Great Lisbon region is explored, showing the usefulness of the proposed scenario reduction method to reduce computational times, and how planning decisions change when health and wellbeing benefits are considered.

--

OR in health services,Long-term care planning,Multi-objective,Stochastic programming,Scenario reduction methods

• Economics and Business - Social Sciences