Designing a Case-Mix Based Reimbursement System for Japanese Long-Term Care Hospital Beds

Posted: 11 Jun 2007

See all articles by Naoki Ikegami

Naoki Ikegami

Keio University - School of Medicine - Department of Health Policy & Management


Background: Long-term care (LTC) hospital beds were reimbursed at a flat per diem rate under Japan's social health insurance which led to the admission of low-care cases and the blurring of differences in function from nursing homes.

Objectives: To design a case-mixed based system for LTC hospital beds that reflects costs and patient characteristics

Methodology: A case-mix grouping matrix, consisting of 3 levels of medical need and 3 levels of ADL, plus additional sub-groups for those with cognitive problems, was first designed. The criteria to place patients into each group were determined from the following. 1) Detailed assessment data and cost data based on the patient's wage-weighted care time and drugs prescribed. These data were collected from 3,538 inpatients in 70 LTC hospitals. 2) Statistical analysis of the relative cost for each of the patient's characteristics and the services provided. 3) Feedback from the hospitals about how appropriately each patient was grouped by the draft criteria according to their perception of the patient's medical acuity and costs, 4) Feedback from expert panel of clinicians, 5) Administrative concerns on up-coding and political pressure from patient groups etc. The survey, analysis and final decisions on the grouping criteria were performed in 2004 to 2006.

Results: The variance explanation of the criteria for wage-weighted care time was 26.9%, which decreased to 24.9% when drug costs were added.

Conclusion: Reimbursement based on the new case-mixed grouping criteria was introduced in July, 2006. However, its intended objectives may only be partially realized because, as a concession to the providers, detailed patient assessment data do not have to be submitted together with the claims, nor are there any mechanisms to monitor the quality of the data or the quality of care. Moreover, although the introduction of the case-mix system was intended to be budget-neutral, as a result of the government's decision to reduce social health insurance expenditures, the volume-weighted average per-diem rate for LTC beds were reduced by about 10% below average costs. These two factors are likely to increase the risk of up-coding. Scientific validity is only one factor in designing and introducing a case-mix based system: the political climate is likely to take precedence. The present criteria have been validated by a new study made in 2006, 5 months after the introduction of the case-mix system. Only very minor changes in the grouping is likely to be made.

Keywords: case-mix based payment, long-term care, up-coding

JEL Classification: I18, H32, H51

Suggested Citation

Ikegami, Naoki, Designing a Case-Mix Based Reimbursement System for Japanese Long-Term Care Hospital Beds. iHEA 2007 6th World Congress: Explorations in Health Economics Paper, Available at SSRN:

Naoki Ikegami (Contact Author)

Keio University - School of Medicine - Department of Health Policy & Management ( email )

35 Shinanomachi
Tokyo 160-8582, Shinjuku-ku
+81-3-5363-3773 (Phone)
+81-3-3225-4828 (Fax)

Here is the Coronavirus
related research on SSRN

Paper statistics

Abstract Views
PlumX Metrics