Visualization of Causal Relationships and Simulation of Medical Expensesof Diabetes Lifestyle Improvement Factors from NDB

2021 Fiscal Year Final Research Report

• Project/Area Number: 18K09934
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 58010:Medical management and medical sociology-related
• Research Institution: Hokkaido University
• Principal Investigator: Ogasawara Katsuhiko 北海道大学, 保健科学研究院, 教授 (90322859)
• Project Period (FY): 2018-04-01 – 2022-03-31
• Keywords: 糖尿病 / 国保レセプト / ベイジアンネットワーク / マルコフモデル

Outline of Final Research Achievements

In recent years, national health care costs have been on the rise. In particular, type 2 diabetes, which is closely related to lifestyle-related diseases, is likely to induce complications. Therefore, it is expected that affected patients will experience high medical costs. Therefore, the objectives of this study are 2 points: (1) To provide effective support for lifestyle modification for diabetic patients in the community. (2) To propose an indicator of the effectiveness of such support. Factors influencing the "intention to improve lifestyle" of diabetic patients and characteristics of regional characteristics were visualized by Bayesian network visualization, and the effect of behavioral change expected by lifestyle improvement was examined. Furthermore, we attempted to conduct a cost-effectiveness analysis using a Markov model with blood glucose control self-management remote appropriation in diabetes mellitus.

Free Research Field

医療情報学、医療システム学

Academic Significance and Societal Importance of the Research Achievements

特定健診データ項目等を用いて地域別ベイジアンネットワークモデルを構築した。「生活習慣の改善意思」に対して、30 分以上の運動習慣とBMI が直接的な影響を与えており、地方部のモデルでは30 分以上の運動習慣と20 歳からの体重変化、性別、飲酒習慣が直接的な影響を与えていることが示された。また，生活習慣の改善によって都市部と地方部共に、1 年間の体重に対して最も影響を与える可能性が示唆された。今後、モデルの妥当性評価を行うことによって、地域性を考慮した健康増進施策の立案および実施への応用が期待できると考えられる。