The Cost Effectiveness Analysis for Taiwan Institutional Care Models in Dementia

Background
Same as many developed countries, the population construction of Taiwan shows an aging societal trend. As the life expectancy becomes longer, we should pay attention to the threats of elder diseases such as chronic diseases. Dementia is a kind of elder diseases. As the elders’ age, the prevalence and severity will keep rising; patients lack the abilities of caring themselves, and caregivers often experience mental or physical burdens. When the family and society cannot find the solution to these problems, the demands of dementia institutional care will sure rise. Nowadays, most of the non-governmental circles practices the “Specialized type” and “Segmented type” institutional care models, but less of the “Mixed type” institutional care models. There are still no relative studies and conclusions found in institutional care models in dementia with both better effectiveness and costs of caring.

Method and Objectives
This study is a prospective study. Applying the methods of purposive sampling and opened-questionnaire to collect north, central and south of Taiwan, 3 models and 19 institutions’ cross-sectional cost data were collected. Dementia residents’ retrospective effectiveness data were collected by chart reviewed. In addition to comparing the average per resident/month care costs of three models, multiple regression was also applied to control residents’ basic characters and degrees of cognitive function and ADL, to predict residents’ changes of scores after one year. Then, the count rate of residents effectiveness changed of these three models were compared, and cost-effectiveness analysis was applied to find out the best Taiwan institutional care models in dementia.

Results
“Specialized type” model is the highest in each resident’s per month care costs in average, “Mixed type” model is the next highest, and the “Segmented type” is the lowest; but there is no statistically significant difference among the three models. No matter what residents’ cognitive function or ADL are, all changes of scores presents the situation of retrogressing in average after one year, and there is statistically significant difference in all models. Comparing the residents’ functional degenerated scores among three different models and to arrange the average degenerated MMSE scores from the lowest to the highest are “Specialized type”, “Mixed type” and “Segmented type” models; to arrange the average degenerated Barthel Index scores from the lowest to the highest are “Mixed type”, “Segmented type” and “Specialized type” models. But there is only statistically significant difference in the average degenerated Barthel Index scores that in “Specialized type” model is higher than in “Mixed type” model. Comparing the predicted residents’ functional degrees one year later, the rate of residents whose degrees of cognitive function gained and kept, the “Specialized type” is the highest, and the “Mixed type” is the lowest; the rate of residents whose degrees of ADL gained and kept that “Segmented type” model is the highest, and the “Specialized type” model is lowest. But this phenomenon is relative to that there are more residents with better ADL in “Specialized type” model, the degenerated scores are also relatively higher because of the natural courses of disease degenerated one year later. In the results of cost-effectiveness analysis, no matter it is on cognitive function or ADL, “Segmented type” model takes the advantage in the cost-effectiveness; through the sensitivity analysis, we find that no matter how manpower costs which takes the largest part of total costs increase or decrease, the results does not change, therefore, the results of this study are stable.

Conclusions
According to the results of cost-effectiveness analysis, “Segmented type” is the better institutional care model in dementia for Taiwan. But when residents’ cognitive functions are mild and moderate, their problematic behaviors might disturb others. Those who lived into the “Segmented type” or “Mixed type” care models were not appropriate; “Specialized type” institutional care model still have its necessity in existing.

Perhaps the costs and effectiveness analytic outcomes in the study are inadequate because of insufficiency of data collection. However, it still should be enough for relative government sections’ reference in distributing resources, and helping long-term care industries and non-governmental circles’ operators to manage themselves appropriately.