[[alternative]]Reliability and validity of the Allen Cognitive Level Screen,6th Edition in patients with schizophrenia

[[abstract]]背景與目的：執行功能損傷為思覺失調症患者認知功能損傷主要的特徵之一。艾倫認知測驗 (Allen Cognitive Level Screen, ACL)為國內外常用於評量思覺失調症患者執行功能評估工具之一。然而，目前尚無研究驗證最新版本艾倫認知測驗第六版 (ACLS-6)於思覺失調症患者之心理計量特性，限制ACLS-6的結果解釋和應用。因此本研究之目的為驗證ACLS-6應用於思覺失調症患者之心理計量特性，包含：施測者內信度、施測者間信度、收斂效度及區辨效度，以確認ACLS-6之實用性及臨床實用價值。方法：本研究募集120位病情穩定之慢性思覺失調症患者，分兩組，每組各60位患者。每位患者進行間隔兩週之二次施測。一組患者由同一施測者進行二次ACLS-6之評量，以驗證施測者內信度。另一組患者之二次ACLS-6評量由不同位施測者進行施測，以驗證施測者間信度。每位患者在第一次施測中，亦接受簡易智能量表 (Mini Mental State Examination, MMSE)、蒙特利爾智能測驗 (Montreal Cognitive Assessment, MoCA)和威斯康辛卡片分類測驗 (Wisconsin Card Sorting Test, WCST)之評量。本研究使用組內相關係數 (intraclass correlation coefficient, ICC)驗證施測者內信度和施測者間信度，以Pearson’s r相關係數計算ACLS-6與MMSE、MoCA和WCST，以及ACLS-6分測驗間之相關程度，以檢驗收斂效度。以獨立樣本t檢定 (independent-sample t-test)檢驗ACLS-6能否區辨不同嚴重程度之思覺失調症患者，以驗證區辨效度。結果：在信度方面，同一施測者或不同施測者測量下，ACLS-6總分ICC值分別為0.77和0.66，同一施測者於七個分測驗(縫紉、瓷片排列、簡單著色、拼圖、菱形著色、花生醬清潔、卡片製作)及不同施測者於五個分測驗(縫紉、瓷片排列、簡單著色、菱形著色、卡片製作)測量之ICC值大於0.40。收斂效度方面，ACLS-6之縫紉、瓷片排列及卡片製作分測驗與MoCA (r=0.35-0.44)、MMSE (r=0.32-0.40)及WCST部分指標 (r=0.33-0.41)有中等相關，拼圖分測驗與MoCA (r=0.33)有中度相關。於ACLS-6分測驗相關性，縫紉分測驗與瓷片排列分測驗的相關性較高 (r=0.87)。區辨效度方面，ACLS-6的四個分測驗(縫紉、瓷片排列、菱形著色、卡片製作)達顯著，可區辨不同嚴重程度之思覺失調症患者。結論：整體而言，ACLS-6應用於思覺失調患者具有中等到好的施測者內信度與施測者間信度，有可接受之收斂效度和區辨效度。ACLS-6九個分測驗運用於慢性思覺失調症患者，縫紉分測驗和瓷片排列分測驗有較好的心理計量特性。關鍵字：執行功能，思覺失調症，艾倫認知測驗，心理計量特性。[[abstract]]Background and purposes: Executive function impairment is one of the main features of cognitive impairment in patients with schizophrenia. The Allen Cognitive Level Screen (ACL) is one of the widely used measure for assessing executive function in patients with schizophrenia. However, no studies verify the psychometric properties of the latest version of the Allen Cognitive Level Screen, 6th edition (ACLS-6) in patients with schizophrenia, which limits the interpretation and application of the ACLS-6 in clinical and research settings. Therefore, the purpose of this study was to verify the psychometric properties of the ACLS-6 in patients with schizophrenia, including intra-rater reliability, inter-rater reliability, convergence validity, and discriminant validity in order to confirm its utility.Method: This study recruited 120 patients with chronic schizophrenia who were stable, divided into two groups of 60 patients each. Each patient was administered the ACLS-6 twice in two weeks. A group of patients was evaluated the ACLS-6 twice by the same rater to verify the intra-rater reliability. The other group of patients was administered the ACLS-6 by the different raters to verify inter-rater reliability. Each patient received the Mini Mental State Examination (MMSE), the Montreal Cognitive Assessment (MoCA), and the Wisconsin Card Sorting Test (WCST) during the first assessment. In this study, the intraclass correlation coefficient (ICC) was used to examine the intra-rater reliability and inter-rater reliability. The Pearson's r was used to estimate correlations among the ACLS-6 MMSE, MoCA and WCST, and correlations among the subtests of the ACLS-6 for examining convergence validity. The independent-sample t-test was analyzed to verify whether the ACLS-6 could distinguish patients with different degrees of severity in the diagnosis of schizophrenia to examine discriminant validity. Results: In terms of reliability, the ICC values of the ACLS-6 total score were 0.77 or 0.66 for the same or different examiners, respectively. For the seven subtests (e.g., Stitching, Copying, Simple Coloring, Frog-Flower-Pot, Diamond Coloring, Peanut Butter, Stencilling) administered by the same examiners and the five subtests (e.g., Stitching, Copying, Simple Coloring, Diamond Coloring, Stencilling) administered by the different examiners, ICC values were greater than 0.40. In terms of convergent validity, moderate correlations were found among the subtests of the ACLS-6 (e.g., Stitching, Copying and Stencilling) and MoCA (r=0.35-0.44), MMSE (r=0.32-0.40) and parts of WCST indices (r=0.33-0.41). The strong correlation was observed between the Copying subtest and the Stitching subtest(r=0.87). In terms of discriminative validity, the four subtests of ACLS-6 (e.g., Stitching, Copying, Diamond Coloring, Stencilling) showed statistically significance, indicating that these subtests could distinguish patients with different degrees of schizophrenia. Conclusion：As a whole, the ACLS-6 has moderate to good inter-tester reliability and intra-tester reliability, acceptable convergence validity and discriminative validity for patients with schizophrenia. In chronic schizophrenia, the nine subtests of ACLS-6 the Stitching and Copying subtest revealed better psychometric properties. Keywords: Executive Functions, Schizophrenia, Allen Cognitive Level Screen, Psychometric properties