@article{oai:niigata-u.repo.nii.ac.jp:00018708, author = {真柄, 彰}, issue = {3}, journal = {新潟医学会雑誌, 新潟医学会雑誌}, month = {Mar}, note = {We tried to predict the difficulty of home care for CVA patients, using Multivariate Statistical Methods. The data were colected by enquete for 289 families caring CVA patients who were discharged after rehabilitation training in our hospital at least one month. We asked 4 items of ADL, 4 items of QOL, 13 items of home care problems, and other social aspects. We also used the data about dementia, aphasia, and apractognosia from the records during hospitalization. We could use 144 cases for the analysis. At the first step we had to grade the home care difficulty by simply adding 12 items of care difficulties. The histogram of the grade (mean 1.0±1.6) showed normal distribution. For the next step we tried to classify the factor items by Principal Component Analysis, and at last we selected 5 factors, which are shown in the below expression for the prediction. By Multiple Regression Analysis, the biggest factor to make difficulty was apractognosia. The biggest factor to reduse difficulty was dementia, the next was transfer ability. Aphasia was neutral. We concluded that we can predict the difficulty of home care by next expression. Difficulty=0.137*Apractognosia+0.05*ADL Ability+0.002*Aphasia-0.02*Transfer Ability-0.03*Dementia The predicted score and the actual score had 0.388 (rS) (n=144) of Spearman's corelation coefficient, that had significant corelation at 1 % error level.}, pages = {95--99}, title = {1)脳卒中のリハビリテーション : 在宅介護における問題点(シンポジウム リハビリテーション医療 : 最近の進歩と問題点, 第484回新潟医学会)}, volume = {109}, year = {1995} }