電腦化數字警醒測驗於中風患者之效度與反應性驗證

Validity and Responsiveness of the Computerized Digit Vigilance Test in patients with stroke

陳姿廷(Tzu-Ting Chen)

指導教授 ： 薛漪平

國立臺灣大學/醫學院/職能治療研究所/碩士(2018年)

https://doi.org/10.6342/NTU201800073

摘要

背景：電腦化數字警醒測驗(computerized Digit Vigilance Test, C-DVT)可方便、有效地評估中風患者之持續性注意力，於慢性中風患者具良好的再測信度、同時效度及生態效度。然而測量結果之有效性易受樣本不同而改變，C-DVT於亞急性中風患者是否具有良好的生態效度、預測效度及反應性仍未知，因此限制此測驗於亞急性中風患者之適用性與臨床及研究之使用價值。
目的：本研究欲驗證C-DVT於亞急性中風患者的效度(含生態效度、預測效度)及反應性(含團體層級、個別層級)，以期確認此測驗之心理計量特性及實用價值。
方法：招募60位亞急性中風患者，於住院二週內完成前測，再間隔一至三週完成後測，二次施測內容相同，包含C-DVT測驗、巴氏量表(Barthel Index, BI)及巴氏量表-補充量表(BI-based Supplementary Scales, BI-SS)的執行能力量表(Ability Scale)。生態效度分析C-DVT與BI、執行能力量表分數前後測結果之相關性；預測效度則分析C-DVT前測與BI、執行能力量表後測分數之相關性。團體層級反應性以配對t檢定(paired t-test)檢驗C-DVT前、後測之平均分數是否達統計顯著差異，再以標準化平均反應值(standardized response mean, SRM)呈現二次測驗平均分數的差異程度；個別層級反應性依據改變分數進退步及是否超過最小可偵測差異值(minimal detectable change, MDC)將患者分成四組，呈現各組人數比例，以驗證C-DVT是否具備個別層級之反應性。後續分析不同施測間隔下(＜14天、≥14天)的反應性，團體層級反應性以配對t檢定二組前、後測之平均分數是否達統計顯著差異，並以SRM呈現差異程度，使用拔靴法(bootstraps)分析二組SRM是否具有顯著差異。個別層級反應性則先計算各組的人數比例後，以邊際同質性檢定(marginal homogeneity test)檢驗不同施測間隔下各組的人數比例是否具有顯著差異，再依據各組人數比例計算其合併標準差(pooled standard deviation)，以z檢定及各組比例95% CI (confidence interval)檢驗相同組別之患者人數比例差異是否顯著。
結果：共49位患者參與此研究(33位男性)，平均年齡56歲。生態效度結果顯示，C-DVT與BI、執行能力量表之前後測分數達中度相關(ρ = 0.43 ~ 0.59)。預測效度方面，C-DVT前測與BI、執行能力量表後測分數達低到中度相關(ρ = 0.39 ~ 0.51)。團體層級反應性分析顯示C-DVT於前、後測之平均分數達顯著差異，SRM為0.31。個別層級反應性之分析結果，32.7%的患者C-DVT呈現進步達顯著程度。分二組之團體層級反應性結果顯示，配對t檢定<14天組C-DVT前後測之平均分數沒有顯著差異，不具有反應性(SRM = -0.05)；≥14天組C-DVT前後測之平均分數呈現顯著差異，且具有中度之效應值(SRM = 0.74)，拔靴法結果顯示<14天與≥14天二組SRM的95% CI不重疊，具有顯著差異。分二組之個別層級反應性結果，<14天、≥14天組分別有21.7%、42.3%患者分數進步達統計顯著，邊際同質性檢定發現<14天、≥14天的二組人數比例分配無顯著差異，比較相同組別人數比例差異也未達統計顯著。
結論：C-DVT於亞急性中風患者具備良好的生態效度及預測效度，可接受的團體層級與個別層級反應性，可幫助臨床及研究人員推論亞急性中風患者ADL之表現與能力、預測出院時ADL功能，以及了解團體及個別患者持續性注意力功能之改變。

關鍵字

中風 ； 持續性注意力 ； 效度 ； 反應性

並列摘要

Background: Computerized Digit Vigilance Test (C-DVT) can be used to assess sustained attention of patients with stroke conveniently and efficiently. The test–retest reliability, concurrent validity and ecological validity of the C-DVT are sufficient in patients with chronic stroke. However, such positive findings can not be assumed for patients with subacute stroke because psychometric properties are population dependent. Thus, the psychometric properties of the C-DVT in patients with subacute stroke remain unknown.
Purposes: To examine validity (including ecological validity and predictive validity) and responsiveness (including group-level responsiveness and individual-level responsiveness) of the C-DVT in patients with subacute stroke.
Methods: We recruited 60 patients with subacute stroke who were admitted to rehabilitation ward. This study contained two assessment sessions with 1to 3-week interval. The first assessment session was completed within two weeks after admission, and the second assessment session was completed whin two weeks before discharge. Subjects were evaluated by the C-DVT, Barthel Index (BI) and BI-based Supplementary Scales (BI-SS)-Ability Scale in both sessions. Researcher examined the ecological validity of C-DVT using Spearman's ρ to analyze the association between the scores of the C-DVT and both the BI and Ability Scale at the same session. The predictive validity using Spearman's ρ to examine the association between the scores of the C-DVT at admission and BI and Ability Scale at discharge. Researcher examined the group-level responsiveness using paired t-test and standardized response mean (SRM). The individual-level responsiveness was examined by the percentages of patients in the 4 categories (significant improvement, nonsignificant improvement, nonsignificant worsening and significant worsening), which was divided into on the basis of the value of minimal detectable change (MDC). In addition, researcher divided patients into 2 groups according to the assessment intervals (<14-days and ≥14-days) for analyzing the responsiveness of the C-DVT. The group-level responsiveness of the 2 groups was examined using paired t-test and SRM, and then researcher compared SRM of the 2 groups using the bootstrap approach. The individual-level responsiveness of the 2 groups was examined by the proportions of patients in the 4 categories. Then, researcher used marginal homogeneity test to examine the distributions of 4 categories of the 2 groups, calculated pooled standard deviation of proportions and used Z-test and proportions 95% CI of each proportions to exmined difference of proportion in the same category of the 2 groups.
Results: 49 patients participated in this study. The C-DVT scores were moderately correlated with the scores of the BI and Ability Scale (ρ = 0.43 ~ 0.59), supporting the ecological validity. Low to moderate association (ρ = 0.39 ~ 0.51) between the scores of the C-DVT at admission and the scores of the BI and Ability Scale at discharge was found, supporting the predictive validity. The group-level responsiveness of C-DVT presented small SRM (0.31) and a significant difference (p = 0.034) between admission and discharge. The individual-level responsiveness of C-DVT showed 32.7% patients in the category of significant improvement. The group-level responsiveness of the <14-days group presented a little SRM (-0.05) and no significant difference, but the ≥14-days group presented moderate SRM (0.74) and a significant difference. The bootstrap analysis showed that SRM of the 2 group was significant difference. The results of individual-level responsiveness of the 2 groups was found 21.7% patients of <14-days group and 42.3% patients of ≥14-days group were in the category of significant improvement. In the 2 groups, the distributions of 4 categories were not significant different. The difference of proportions in the same category of the 2 groups was not
statistically significant.
Conclusion: The C-DVT has good ecological validity and predictive validity, and acceptable group-level responsiveness and individual-level responsiveness in patients with subacute stroke. These findings support the result of C-DVT can be inferred the patient’s ADL performance and ability, and predicted patient’s ADL function at discharge. The C-DVT also can be an ideal outcome measure to understand the changes of patient’s sustained attention for clinical professionals and researchers.

並列關鍵字

stroke ； sustained attention ； validity ； responsiveness

參考文獻

陳蕙君（2008）。日常注意力測驗應用於中風個案之心理計量特性驗證（碩士論文）。取自取自華藝線上圖書館。（系統編號U0001-2307200812483700）

王乙婷、何美慧（2003）。自我教導策略增進ADHD兒童持續性注意力之效果。特殊教育學報，18，21-54。

林鋐宇、張文典、洪福源（2011）。注意力的神經生理機制。身心障礙研究季刊，9（2），123-134。

Hsueh, I. P., Mao, H. F., Huang, H. L., & Hsieh, C. L. (2001). Clinical applications of balance measures in stroke patients. Formosan Journal of Medicine, 5(3), 261-268. doi: 10.6320/FJM.2001.5(3).02

喬慧燕、薛漪平、古佳苓、陳蕙君、謝清麟(一)2015）。常用於中風研究的五種注意力測驗之心理計量特性比較。台灣復健醫學雜誌，35（2），65-72。doi: 10.6315/2007.35(2)01