INTRODUCTION
The long-term care insurance (LTCI) system has been implemented in a few countries including Korea, Japan, and Germany [
1,
2]. The LTCI is a system to improve the quality of life of the elderly and to contribute to the family's welfare by providing care benefits to elderly persons who cannot maintain the activities of daily living (ADL) without assistance [
3]. The policy consists of providing social services that complement the care given by families rather than providing alternative medical services. The LTCI in Korea covers senior citizens who are 65 years old or older and those who are less than 65 years old and suffering from geriatric disease [
4].
For acceptance into the LTCI program, an evaluation agent, such as a nurse, social worker or physical therapist, visits the home of the person applying for long-term care for evaluation of the condition of the applicant, using 5 evaluation domains [
3,
5] (
Fig. 1). The long-term care grading committee reviews the status of the applicant and doctor's medical opinion to decide the care grade the applicant should receive. The Korean LTCI employs a system of 3 care grades. Elderly people who have to depend on assistance in all aspects of daily life are categorized into care grade 1. Senior citizens who need continuous instruction and monitoring are classified into care grade 2. Lastly, aged persons who need some assistance when going out are categorized into care grade 3.
Since the introduction of the Korean LTCI in 2008, the beneficiaries of this social insurance scheme have numbered up to 323,000 elderly persons in 2012. Because doctors are not able to evaluate all of the applicants, a valid evaluation system with qualified agents is mandatory. To assess the validities of the 5 domains of the evaluation system, the Korea Institute for Health and Social Affairs (KIHASA) collected data on the elderly who had been using long-term care services in the period from August 2011 through September 2011 [
5].
The evaluation tool for measuring motor impairment of the rehabilitation domain, named the Motor Impairment Scale (MIS), was introduced and adopted despite insufficient evidence for its validity and reliability. Although the MIS has been used by up to 950,000 applicants since 2008 as part of a motor evaluation tool in LTCI, there have been few reports examining the validity and reliability of the MIS. The aim of the present study was to investigate the relationships of the MIS with service time for ADL, ADL score and care grade by using objective data from the KIHASA. To the best of the authors' knowledge, this is the first study to report the validity of the MIS in the Korean LTCI system.
DISCUSSION
There have been many studies concerning the adequacy of evaluation items in the LTCI in accurately assessing the care needs of the elderly [
16-
19]. However, the studies on the validity and reliability of the evaluation tools in the LTCI system have been rather limited up until now [
20,
21]. A study revealed that care grade and cognitive impairment were generally correlated, but some adjustment measure for cognitive impairment was needed in mildly or moderately physically disabled patients [
6]. Another study reported that physical function classified by the LTCI system of Japan was correlated with Fried's criteria for frailty syndrome [
22].
In the process of long-term care grade judgment, an evaluation agent assesses the applicant according to the 5 domains of physical functions, which include cognition, behavioral changes, demand for nursing, and demand for rehabilitation. Among the 5 domains, the rehabilitation domain consists of MIS and joint range of motion limitation. The present study was designed to investigate the validity of the MIS in the LTCI system of Korea. The service time offered for ADL was considered a primary criterion measurement to verify the concurrent validity of the MIS because the payments of LTCI are based on the estimated service time. The ADL evaluations were used as a secondary criterion measurement and not as a primary criterion measurement because the validities of ADL items and score are still unproven. Care grade was used only as a supportive criterion measurement, since it had been assigned before this study. We considered the possibility that the applicants might give false information to the evaluation agents to obtain a higher care grade. The objective of the current study was achieved by demonstrating a significant correlation between MIS and service time for ADL and ADL evaluations.
The characteristics of FG and DG were quite different, resulting in different patterns of correlations. The total ADL score, cognition score, service time, and sex of FG and DG were significantly different. The distribution of care grade was dissimilar between FG and DG, and care grade 3 comprised a larger proportion in DG. The elderly in FG showed significantly more dependent ADL and lower cognition scores than those in DG, suggesting more disability among the elderly in FG. Care staff members spent more time on ADL in FG than in DG. The mean age of total participants was approximately equal to the life expectancy of Korean males (77.2 years in 2010), but lower than that of females (84.1 years in 2010). We inferred that the predominance of females in both groups was caused by the longer life expectancy of women [
23]. Because there was no statistical difference of MIS between FG and DG, the baseline differences of service time and total ADL score between both groups were assumed to have originated from other factors, such as cognition-related factors or medical disease [
8].
Service time for ADL was our primary criterion measurement to gauge the concurrent validity of MIS, because the estimated care time is an important determinant of care need and care grade in Korea and Japan [
5,
13]. In FG, significant correlation was observed between MIS and service time. U-MIS was an independent predictor after adjusting for age, sex, and cognition score in FG. After adjustment for other factors, L-MIS was not an independent predictor for service time. The motor function of the upper limbs may have more influence on the performance of ADL activities than that of lower limbs. The correlation between MIS and service time in DG was different from our prediction. No relationship between MIS and service time was demonstrated in DG, and only 2.4% of the variation in service time could be explained by age, sex, U-MIS, and L-MIS. Considering these results, our postulation that the service time offered for aiding ADL might be representative of the care needs can be applicable to FG only.
ADL was a secondary criterion measurement to demonstrate the validity of MIS. It was proved that the MIS was correlated with all of the ADL items and the total ADL score. Most of those correlations were moderate except for the weak association of MIS and bathing and bladder control in DG. FG showed stronger correlation between MIS and ADL than DG. After adjustment of age, sex, and cognitive score, U-MIS and L-MIS were independent predictors for the total ADL score in both groups. In conclusion, as the motor functions of the elderly subjects were more impaired, the ADL of these subjects were more dependent. In addition, living alone was another independent predictor for the total ADL score in DG. The elderly subjects who lived alone led a more independent daily life.
The relationships of care grade with MIS, service time, and total ADL score were analyzed to find characteristics of service time in DG, because no correlation was found between MIS and service time. Significant correlations between care grade and MIS, service time, and ADL were demonstrated in FG. In DG, care grade showed a weaker correlation with MIS and total ADL score than in FG. There was no correlation between care grade and service time for ADL. Service time of DG showed no correlation with any of the other variables including MIS and total ADL score. Therefore, service time offered for ADL cannot be used to estimate the ADL and care grade in DG.
A possible explanation for this finding is that the service offered by staff members in DG may not be related to the ADL. There is research showing that the use of major services in domiciliary elderly care was decided more by the needs of the caregivers than by the care grades of the applicants, suggesting that consideration of the caregiver situation should be included in policy making [
24]. It is obvious that the services offered by care staff members should be varied, considering the diverse life environments of domiciliary senior citizens. The services can be social such as communication and even just supportive conversation, or related with instrumental ADL such as shopping and banking rather than ADL. In these situations, the care staff members offer special services, while the caregivers help with daily activities. Further research to reveal the service types that the caregivers need in reality are required. A more upgraded system which incorporates individual home environments and provides more suitable services on a case by case basis should be introduced.
Some researchers have sought to develop a simple method of estimation using the ADL category to predict care grade [
10]. The accuracy rate for the estimation of care grade by care after urination, walking and eating was 66.7% in the physically disabled facility-care elderly. The current study proved that the MIS was an independent predictor for the service time and the total ADL score. The importance of MIS as an evaluation tool comes from its objective nature compared to the evaluation tools of other domains including ADL and cognition. When the applicants are evaluated by agents, most of the domains are assessed by interviewing the applicants themselves and their caregivers. The domain for rehabilitation demands, composed of motor impairment and joint limitation, is crucial for objective evaluation because it is the only domain in which the agents observe the actual performance of the applicants. Accordingly, in the cases that the caregivers are not cooperative enough to be interviewed or complete the questionnaire, we can estimate the ADL level by simply assessing the motor impairment and cognition score of the elderly. In addition, some malingerers who describe their status as worse than it is can be detected during motor evaluation if they have good cognitive function.
The main limitation of the current study is the exclusion of dementia patients. It has been found that patients with dementia and diseases of the circulatory system, especially cerebrovascular disease, are the most common recipients of care in the LTCI system [
1]. The current study can only be applied to the physically disabled elderly because dementia patients were excluded in the study design. Therefore, there is no validity evidence for the other major group of recipients, dementia patients. Further research investigating the validity of MIS in the elderly with dementia is necessary. In addition, because only the criterion validity was proved in the present study, future studies are required to ascertain the reliability of MIS as a useful tool in LTCI.
In conclusion, we found that the MIS had significant validity in predicting service time and ADL in the elderly admitted to a facility. The validity of the MIS was only partially proved in the domiciliary elderly because MIS correlated only with ADL and not with service time for ADL in these participants. Its short measurement time, the ease of learning by evaluation agents, and the correlation with service time and ADL makes the MIS a useful evaluation tool in the LTCI system of Korea. Future studies on the validity of the MIS in dementia patients as well as studies on the reliability of the MIS are necessary.