Objective : Local authorities face real challenges when it comes to annual budget planning for funding the system of long-term care. Uncertainty about the long-term cost of caring for current residents in the system, in addition to unknown future admissions, have made the tasks of local authority budget managers very complex and demanding. In this paper, we present a software implementation of a novel forecasting framework developed by the authors to provide useful information to local authority budget planners involved in long-term care. Methods : The tool is built upon a forecasting framework, which combines unit costs of care with an estimated underlying survival model for publicly funded residents in long-term care, to provide forecasts of the cost of maintaining the group of elderly people who are currently in long-term care (referred to as known commitments) for a period of time. A prototype version of the software tool, which was created and tested in collaboration with an English borough, allows user interaction via a friendly graphical interface that guides through a set of screens of options in a familiar wizard fashion. Results and discussion : Feedback from care planners and managers show that the tool helps them gain better understanding of the patterns of length-of-stay of residents under their care, and provides quantitative inputs into their decision making on budget planning for long-term care. The development of the software tool brings advanced modelling techniques out of research papers into the hands of decision makers in the public sector and contributes to improving the delivery of long-term care. |