| Development of Global Deterioration Scale Staging Algorithm |
|
Chang Won Won, Jong Woo Paik, Key Chung Park, Sang Yun Kim, Kun Woo Park, Dong Woo Lee, Sang Tae Han |
1Department of Family Medicine, Kyung Hee University School of Medicine, Seoul, Korea. chunwon62@dreamwiz.com
2East-West Medical Research Institute, Kyung Hee University, Seoul, Korea.
3Department of Psychiatry, Kyung Hee University School of Medicine, Seoul, Korea.
4Department of Neurology, Kyung Hee University School of Medicine, Seoul, Korea.
5Department of Neurology, Seoul National University Bundang Hospital, Seoul National University College of Medicine, Seongnam, Korea.
6Department of Neurology, Korea University College of Medicine, Seoul, Korea.
7Department of Psychiatry, Inje University Sanggye Paik Hospital, Seoul, Korea.
8Department of Informational Statistics, Hoseo University College of Natural Science, Asan, Korea. |
| 치매의 등급척도인 전반적퇴화척도(Global Deterioration Scale)의 등급판정 알고리듬 개발 및 타당성 조사 |
|
원장원1,2,백종우3,박기정4,김상윤5,박건우6,이동우7,한상태8 |
|
|
|
|
| Abstract |
BACKGROUND
The Global Deterioration Scale (GDS) is a useful tool for staging dementia; each stage is described by specific characteristics. However, one should not rely on the presence or absence of a single symptom in determining the stage. There is a need for a systematic computerized program to enable untrained doctors to easily assess dementia. This study aimed to generate an algorithm to help stage dementia.
METHODS
Items were drawn from each stage and sorted out into questions adequate for the caregiver and questions adequate for the patient. Subjects recruited were 50 years or older and had visited the neurologic and/or psychiatric clinic at any of the university affiliated hospitals with symptoms of memory impairment. Structured questionnaires with 20 questions were administered to the subject-informant dyads. Psychometricians or well-trained nurses then assessed the remaining 10 items and decided the overall stage. Classification tree analysis was accomplished by using SPSS Answer Tree 3.0 software.
RESULTS
182 subject-informant dyads were included in the analysis. The mean age was 74.5 years; 112 (61.5%) were female. Among the 30 predictors, the item 'get lost when travelling' was the most important predictor of GDS score (chi2=96.6, p=0.0000). The classification tree algorithm begins with the item 'get lost when travelling' and includes 13 predicting variables. The most probable GDS predicted scores are presented in the final nodes of the algorithm. Risk estimate, probability of misclassification in the developed model, was 35.2%.
CONCLUSION
A classification tree algorithm for GDS staging was developed to narrow down the range of choices when staging cognitive impairment. The algorithm is yet to undergo validity and reliability tests. |
| Key Words:
Global Deterioration Scale, Algorithms, Dementia |
|
PDF Links
Full text via DOI
Download Citation
