A scoring system and seven factors associated with certification for Japanese long-term care insurance in older people

Abstract

Introduction: The increase in the older population is a serious concern in developed countries, and how to maintain independence of these individuals is now an urgent issue. Various factors are known to put older people at risk for needing long-term care, but it is not clear to what extent each factor is associated with that need.

Materials and methods: In a cohort of 1577 community-dwelling older persons, we excluded 40 persons whose long-term care insurance certification was unknown and then divided the remaining 1537 into two groups: dependent group (134 persons) certified as requiring assistance or long-term care, and an independent group (1403 persons). We extracted 7 factors and created a scoring system from these factors based on regression coefficients.

Results: Among 92 factors initially evaluated, 7 were significantly associated with the need for assistance or long-term care, namely walking speed, age, grip strength, mobility (EQ5D), ability to use public transportation by oneself (IADL), ability to perform usual activities (EQ5D), and serum albumin levels. Based on these 7, we constructed a scoring system and calculated a cutoff value of 8 points with an area under curve as high as 0.949.

Conclusion: We determined the cutoff value for dependency risk to be 8, but no single factor scored 8 or higher, suggesting that a combination of these factors promotes the need for nursing care in older people.

Keywords: Long-term care insurance (LTCI); Older adults; Scoring system.

Fig. 1

Participants included in the study. We enrolled 1577 community-dwelling older persons. Forty were excluded as their eligibility for assistance or long-term care insurance was unknown at the time of entry. The remaining 1537 participants were included

Fig. 2

Receiver operating characteristic curve for factors significantly associated with assistance or long-term care certification. We calculated the area under the curve (AUC) for factors significantly associated with assistance or long-term care certification, namely, Age, Barthel index score, IADL score, EQ5D score, walking speed, grip strength, and MMSE score, and determined it to be 0.946

Fig. 3

Cutoff score for factors significantly associated with assistance or long-term care certification. We examined the cutoff score for factors significantly associated with assistance or long-term care certification by evaluating the diagnostic accuracy of quantitative tests using ROC curves. The cutoff score calculated based on the sum of sensitivity and specificity was determined to be 8 with a sensitivity of 0.903, specificity of 0.890, and an AUC of 0.946 (95% CI 0.926–0.966)

Fig. 4

Total score distributions of participants in the independent group. Histogram analysis showing distribution of total scores of 1,403 individuals in the independent group. The number of individuals scoring 8 or higher was 155 out of 1403

Fig. 5

Total score distributions of participants in the dependent group. Histogram analysis showing distribution of total scores of 134 individuals in the dependent group. The number of individuals scoring 8 or higher was 121 out of 134