[Long-term differences in cumulative long-term care costs according to risk assessment scale scores for functional...

Objectives　A previous study reported that future long-term care (LTC) costs differ according to risk assessment scores for incident functional disability. This study uses extended long-term follow-up data to examine cumulative LTC cost differences across various follow-up periods.Methods　As part of the Japan Gerontological Evaluation Study (JAGES), a baseline survey was conducted in August 2010 and January 2011 among individuals aged ≥ 65 years in five Japanese municipalities (response rate: 63.7%). These individuals were not certified as requiring public LTC insurance benefits. Using public LTC claim records, LTC service costs were tracked for 21,626 individuals aged over 12 years following the baseline survey. The risk assessment scale for incident functional disability (0-48) comprises 12 items including sex and age. An ordinary linear regression model, Tobit regression model, and a linear regression model with multiple imputation for missing values were applied.Results　During the follow-up period, 9,368 (43.3%) participants were certified as requiring LTC or support, and 7,137 (33.0%) died. Even at the 12-year follow-up, the higher-risk group (risk score ≥17) was positively associated with the cumulative cost of LTC services per person, duration of LTC service use, and the proportion of people certified as requiring LTC or support. Compared with the difference between high- and low-scoring groups at the 1-year follow-up, the magnitude of the difference increased exponentially: a 59.7-fold increase at 6 years, a 120.2-fold increase at 9 years, and a 185.6-fold increase at 12 years. After adjusting for confounders, the 12-year cumulative LTC cost increased by approximately 80,900 yen (95% confidence interval [CI]: 68.6-93.2) per 1-point increase in the risk score. The corresponding estimates were 63,000 yen (95%CI: 49.3-76.6) in the low-score group (risk score ≤16) and 110,400 yen (95%CI: 80.6-140.2) in the high-score group (risk score ≥17). Total expenditure increased with longer observation periods, although the increase plateaued after approximately 10 years in the high-score group. The major results and trends were generally consistent across the different model specifications.Conclusion　The baseline risk assessment scale for incident functional disability can be used to estimate the subsequent occurrence of LTC certification and cumulative LTC expenditure over a 12-year follow-up.