Application Of Linear Regression Analysis Of Length Of Hospital Stay For Patients With Acute Ischemic Stroke—Experience From A Teaching Hospital At West Coast Area In Middle Taiwan

• Main Authors: Lin Ching Mao, 林青茂
• Other Authors: Weng Shao-Jen
• Format: Others
• Language: zh-TW
• Published: 2012
• Online Access: http://ndltd.ncl.edu.tw/handle/68570000432949805267

碩士 === 東海大學 === 工業工程與經營資訊學系 === 100 === Background: Stroke will be a major impact on national productivity and household economy. In order to provide the references of medical staff or family clinical care. This study investigate correlation between length of hospital stay in patient with acute ischemic stroke and medical variables. Materials&Methods: This is a cross-sectional study. I gathered 213 patients all over 40 years old from a teaching hospital at west coast area in middle Taiwan. Analysis of medical records data by using Minitab and SPSS softwares. Results: All variables including age, sex, body mass index, NIHSS score,diseases history and lifestyles illustrated no positive correlation and were independent each other.Best subsets linear regression model showed age,body mass index,NIHSS score and cardiovascular diseases significantly influences length of hospital stay, but plot of residuals versus the fitted values show an outward-opening picture and indicated that non-normal distribution found. I used the Box-Cox transformation for the response variables of length of hospital stay and new results indicated that the variance is stable and corresponding ANOVA test got statistical significance. Patients with acute ischemic stroke in this study, 55 victims had length of hospital stay≧14 days and 158 victims had length of hospital stay＜14 days. Application of simple logistic regression analysis showed age（odds ratio 1.051, 95％ confidence interval 1.021-1.080）and NIHSS score（odds ratio 1.146, 95％ confidence interval 1.088-1.207） both influencing length of hospital stay ≧14 days in ischemic stroke victims. Chi-squared test discovered that age influencing length of hospital stay ≧14 days in patients with acute ischemic stroke（odds ratio 2.500, 95％ confidence interval 1.284-4.891）and increases in the NIHSS score indicated higher probability of length of hospital stay ≧14 days which all mean statistical significance. The research findings provide the National Health Insurance Bureau for developing Two-diagnosis related groups of ischemic stroke payment system references.