Read more

Using big data analysis in developing the future of long-term care

Among the various challenges currently facing Japan, long-term care is one which is particularly pronounced. Japanese society is rapidly aging before our eyes, with one in three people being 65 years of age or older by 2025. As a result, Japan is confronted with the problem of ensuring that there are enough carers and of somehow controlling rising long-term care costs.

It is in this context that the researchers in the Sugawara Laboratory of the School of Management’s Department of Business Economics are applying economic analysis to Japanese long-term care. The question they are focused on is, given the limited financial resources for social welfare and social security, what is the best approach to long-term care that will ensure everyone has good quality of life? These researchers are looking at data to see how funds for long-term care are being used, how they can be used more effectively, etc., in order to provide a theoretical backing for different policies.

To conduct their analysis, the researchers are looking at “long-term care receipt data.” That is, they are looking at the data related to payments made for long-term care services, i.e., the data obtained from long-term care “receipts.” 
This data, however, is not data that was gathered for the purposes of analysis, it is simply raw, collected data, and it is in extracting useful data from this that the researchers show their skill. In order to successfully extract meaningful data, the Sugawara Laboratory researchers made use of pattern recognition-focused machine learning.

Looking at data to obtain a “ground-level” view of the actual situation

The collected long-term care receipt data contains data from roughly 80% of those who are using long-term care insurance. This amassed big data set encompassing roughly 5 million people over a 120 month period took around three days just to organize. Handling this much data is an immense task, but, with patience and perseverance, researchers were able to obtain a number of revealing insights from it.

It is those working “on-the-ground” in long-term care who are most familiar with the realities of the system, but it is often hard for them to get their voices heard. All the more reason, therefore, the researchers in the Sugawara Laboratory believe, to use data analysis to highlight facts that those on the periphery of the long-term care system will recognize as problems and will then work on developing specific solutions to address.

Taking action to resolve the challenges of long-term care overlaps with goal #3 of the global SDGs: “Ensure healthy lives and promote well-being for all at all ages.” Looking ahead, it is hoped that this data-driven approach will contribute to a proper assessment of the long-term care system and its performance which, in turn, can be used to inform future policy recommendations.

School of Management Department of Business Economics
Associate Professor Shinya SUGAWARA

■ Main research content

Fields of research encompass economic analysis of aged care. Researchers incorporate Bayesian statistics, machine learning and other data science approaches in conducting multi-faceted analysis of the long-term care challenges facing an aging Japanese society.

Recommended Initiatives