ANALYSIS OF THE RELATIONSHIP BETWEEN PHYSICIAN DENSITY AND HEALTHCARE EXPENDITURE IN JAPAN:USING LARGE-SCALE REAL-WORLD DATA OF EMPLOYMENT-BASED HEALTH INSURANCE CLAIMS (WELLNESS STAR DATABASE)
Author(s)
Akihiro Kakinuma, BS1, Takao Yabuki, M.S.1, Yuri Miyamori, M.Eng1, Masaru Kinugawa, M.S.1, Tomomi Takeshima, PhD1, Makoto Takamoto, M.Econ.1, Kosuke Iwasaki, MBA2, Ayano Chida, High School Diploma2, Naoki Ikegami, PhD2, Ataru Igarashi, PhD3.
1Nippon Life Insurance Company, Minato-ku, Japan, 2Milliman, Inc., Tokyo, Japan, 3Department of Health Economics & Outcomes Research, Faculty of Pharmacy, Tokyo University, Tokyo, Japan.
1Nippon Life Insurance Company, Minato-ku, Japan, 2Milliman, Inc., Tokyo, Japan, 3Department of Health Economics & Outcomes Research, Faculty of Pharmacy, Tokyo University, Tokyo, Japan.
OBJECTIVES: Physician-induced demand (PID) refers to the provision or recommendation of medical services by physicians that exceed what patients would choose if fully informed. In Japan, physician density is positively associated with healthcare expenditure; however, it is unclear whether this reflects true PID or lower access costs in high-density areas. This study examined the existence of “true” physician-induced demand in Japan using individual-level health insurance claims data.
METHODS: We analyzed the monthly expenditures for each individual, aggregated from claim data in the Wellness Star database owned by Nippon Life Insurance, comprising 251 million beneficiary-months from more than 200 employment-based health insurance societies. The analysis focused on fiscal year 2022. Industry groups with similar demographic-adjusted per-member-per-month (PMPM) expenditures were selected. Within these groups, insurers in prefectures with high physician density (H group) and low physician density (L group), based on Ministry of Health, Labor, and Welfare statistics, were identified. Disease-specific prevalence, expenditure per patient per month (PPPM), PMPM expenditure per insured person, and service frequencies were compared.
RESULTS: The L group included 468,920 person-months from Gifu, Ibaraki, and Shizuoka, while the H group included 274,624 person-months from Fukuoka, Kochi, Tokushima, and Okayama. The largest difference was observed among men aged 45-54 years, with PMPM expenditure 1.5 times higher in the H group. The difference was most pronounced for gastroesophageal reflux disease (B/G ratio 1.7). Decomposition showed no difference in prevalence but a substantial difference in PPPM (ratio 1.6).
CONCLUSIONS: Higher physician density was associated with higher healthcare expenditure, driven primarily by higher expenditure per patient. These findings may suggest the presence of physician-induced demand.
METHODS: We analyzed the monthly expenditures for each individual, aggregated from claim data in the Wellness Star database owned by Nippon Life Insurance, comprising 251 million beneficiary-months from more than 200 employment-based health insurance societies. The analysis focused on fiscal year 2022. Industry groups with similar demographic-adjusted per-member-per-month (PMPM) expenditures were selected. Within these groups, insurers in prefectures with high physician density (H group) and low physician density (L group), based on Ministry of Health, Labor, and Welfare statistics, were identified. Disease-specific prevalence, expenditure per patient per month (PPPM), PMPM expenditure per insured person, and service frequencies were compared.
RESULTS: The L group included 468,920 person-months from Gifu, Ibaraki, and Shizuoka, while the H group included 274,624 person-months from Fukuoka, Kochi, Tokushima, and Okayama. The largest difference was observed among men aged 45-54 years, with PMPM expenditure 1.5 times higher in the H group. The difference was most pronounced for gastroesophageal reflux disease (B/G ratio 1.7). Decomposition showed no difference in prevalence but a substantial difference in PPPM (ratio 1.6).
CONCLUSIONS: Higher physician density was associated with higher healthcare expenditure, driven primarily by higher expenditure per patient. These findings may suggest the presence of physician-induced demand.
Conference/Value in Health Info
2026-05, ISPOR 2026, Philadelphia, PA, USA
Value in Health, Volume 29, Issue S6
Code
EE196
Topic
Economic Evaluation
Topic Subcategory
Cost/Cost of Illness/Resource Use Studies
Disease
No Additional Disease & Conditions/Specialized Treatment Areas