An analysis of Blood Pressure and Medical Expenditures using the Medical Checkup and Receipt Dataset

Author Name NAWATA Kazumitsu (Faculty Fellow, RIETI) / MATSUMOTO Akikuni (University of Tokyo) / KIMURA Moriyo (The Public Health Institute)
Creation Date/NO. May 2019 19-J-031
Research Project Exploring Inhibition of Medical Expenditure Expansion and Health-oriented Business Management Based on Evidence-based Medicine
Download / Links


Hypertension is considered as one of the most important worldwide health risks and many studies have been done about hypertension. In November 2017, The American College of Cardiology (ACC), American Heart Association (AHA) and nine other organizations announced a new hypertension guideline (2017 ACC/AHA guideline). According to the 2017 ACC/AHA guideline, the criterion of hypertension has changed to 130/80 mmHg from the previous 140/90 mmHg.

In this paper, we analyzed the relationship between blood pressure (BP) and medical expenditures. First, we made a database containing 175,123 medical checkups and 6,312,125 receipts from 88,211 individuals obtained from three health insurance societies. The sample period was fiscal year 2013 to fiscal year 2016. We then evaluated the distributions of BP and factors affecting BP using regression models. We found out that age, gender, height, BMI (body mass index) and some lifestyles affected BP. Next, we analyzed the relationship between BP and medical expenditures using power transformation Tobit models. Although there was a positive relationship between systolic BP (SBP) and medical expenditures found with a simple two-variable analysis, we observed negative effects of SBP on the medical expenditures in the power transformation Tobit models. When age, gender and BMI were included in the model, the estimate of SBP became negative, indicating a high importance of considering the relationships between covariates. Finally, we discussed the sample selection bias and time-varying variables in the Cox proportional hazard model as the problems of previous studies. The results of this study did not support the new 2017 ACC/AHA guideline, at least for SBP. A wide and careful range of reviews not only for heart diseases but also for other disease types should be done. New studies to verify the guideline will also be absolutely necessary.