常用於中風研究的五種注意力測驗之心理計量特性比較

A Review of Psychometric Properties of Five Tests of Attention Commonly Used in Patients with Stroke

喬慧燕(Hui-Yan Chiau) ； 薛漪平(I-Ping Hsueh) ； 古佳苓(Chia-Lin Koh) ； 陳蕙君(Hui-Chun Chen) ； 謝清麟(Chin-Ling Hsieh)

《台灣復健醫學雜誌》 35卷2期 (2007/06) Pp. 65-72

https://doi.org/10.6315/2007.35(2)01

摘要

研究背景與目的：注意力受損是中風患者常見的問題之一。注意力為認知活動如記憶、學習之基礎，若患者之注意力不足，可能影響其復健成效及功能恢復。故精確評量患者之注意力功能，方可擬定適當之治療計畫與介入方式。精確評量之首要關鍵，在於使用心理計量特性良好之注意力測驗。然而，目前用於中風患者之注意力測驗的心理計量特性仍不清楚，故本研究彙整2003-2005年間常用於中風實證研究之注意力測驗，比較其用於中風患者之心理計量特性（含信度、效度及反應性），以作為臨床工作與研究者選擇注意力測驗之參考依據，並提供後續研究之建議。方法：研究者合併檢索電子期刊資料庫（MEDLINE, PsycINFO, CINAHL, 中華民國期刊論文索引系統），找出2003-2005年間較常用（被使用2次以上）於中風研究之注意力測驗，再比較這些測驗應用於中風患者之心理計量特性。結果：常用於中風研究之注意力測驗共計5種：路徑描繪測驗(Trail Making Test, TMT)、認知藥物研究電腦化測驗(Cognitive Drug Research computerized battery)、日常注意力測驗(Test of Everyday Attention, TEA)、數字廣度測驗(Digit Span Task)及Stroop色字測驗(Stroop Color Word Test)。這些測驗針對中風患者之心理計量特性驗證結果如下：一、信度：僅TEA之部分子測驗再測信度良好，其餘工具欠缺信度驗證；二、效度：5種注意力測驗皆無同時效度資料。預測效度僅TMT曾被檢驗，其預測效度在可接受程度。收斂效度僅TEA有相關的驗證資料，但其結果不甚理想。三、反應性：尚未有研究驗證此5種注意力測驗應用於中風患者之反應性。綜言之，此5種常用於中風研究之注意力測驗，鮮少具有針對中風患者之心理計量特性驗證，僅TEA與TMT具備部分心理計量特性驗證，然而，二者之證據亦不完整，故難以比較。結論：本回顧結果顯示常用於中風研究之五種注意力測驗欠缺完整之心理計量特性驗證，使得其用於中風患者之測驗結果精確與否缺乏實證。建議未來研究應驗證此5 種注意力測驗用於中風患者之信度、效度及反應性，以建立測驗用於中風患者注意力評量之有效性證據，供中風復健專業之臨床與學術研究使用。

關鍵字

中風 ； 認知 ； 注意力 ； 心理計量特性

並列摘要

Background and purpose: Comparison of the psychometric properties of tests of attention can provide guidelines with which clinicians and researchers can determine a useful measurement. The purpose of this study is to review and appraise the psychometric properties of attentional tests frequently used during the last 3 years in empirical studies of stroke and to determine how well these tests can be applied on people with stroke.
Methods: This study included two parts: first, a search from the MEDLINE, PsycINFO, CINAHL and Index to Chinese Periodical Literature databases for attentional tests used more than twice in empirical studies of stroke in the period spanning 2003-2005; and second, a search for validation studies of those selected attentional tests to appraise their psychometric properties, including their reliability, validity, and responsiveness, to determine the strengths and weaknesses of existing attentional tests.
Results: Five frequently (i.e. used more than twice) used attentional tests were identified: the Trail Making Test (TMT), the Cognitive Drug Research computerized battery, the Test of Everyday Attention (TEA), the Digit Span Task, and the Stroop Color Word Test. We found that the reliability of these tests had not been examined in stroke patients except for the TEA, which has adequate to good reliability. The validity of these tests had not been examined completely. We found that the TMT had acceptable predictive validity for on-road driving performance and that the convergent validity of the TEA was not satisfactory. However, there were no validation studies for the concurrent validity of these tests as applied to stroke patients. In addition, the responsiveness of the five tests has not been examined in patients with stroke.
Conclusion: Most of the psychometric properties of these five tests of attention have been rarely examined. Further validation studies of these five attention tests in stroke patients are warranted in order to justify the usage of these five tests.

並列關鍵字

stroke ； cognition ； attention ； psychometric properties