OBJECTIVES:

Japan has a compulsory insurance system ensuring citizens the universal access to affordable healthcare. Although the government runs the National Database incorporating all public health insurance systems, its poor accessibility eludes researchers. Alternative commercial claims databases have limitations in their generalizability. This research, using major commercial claims databases, assessed the characteristics of major health insurance systems; employee-based health insurance (EHI) for employees in mid-large companies and their families and the citizens’ health insurance (NHI) for non-/self-employed workers.

METHODS:

The JMDC database covering EHI data (2005–2021; JMDC Inc.) and DeSC database covering EHI and NHI data (2014–2021; DeSC Healthcare, Inc.) were used. Both in EHI and NHI, age- and gender-specific medical costs per-member-per-month were calculated in subgroups by Charlson Comorbidity Index (CCI; CCI=0, CCI>0) and 15 disease categories of ICD-10. Age- and gender-specific prevalence was analyzed for different diseases.

RESULTS:

During the study period, 13,665,051 patients were in JMDC, 908,237 in DeSC-EHI, and 7,147,428 in DeSC-NHI. In 2020, the overall medical cost was higher in DeSC-NHI. The same trend was observed even under the stratified analysis with CCI. The medical cost for each disease was especially higher in DeSC-NHI for male aged >20 years (≥1.5 fold). Prevalence rates were comparable in most disease categories, while those of DeSC-NHI were 3-8% higher in both genders aged ≥30 years in F00-F99 (mental) and G00-G99 (nervous system); prevalence rates of EHI were 3-13% higher in both genders in H00-H59 (eye). The costs were higher in DeSC-NHI especially in adult male, and the prevalence was different in certain diseases categories.

CONCLUSIONS:

Socioeconomic factors may contribute to the characteristics of the two claims databases on the major health insurance systems. It is important for researchers to note that health insurance systems have different characteristics to develop research plans or interpret the data.