影響認知障礙症患者 MoCA-T量表命名項目作答的可能因素之探究

中文摘要研究背景與動機：認知障礙症為漸進性退化疾病，隨高齡化加劇而快速增長，對醫療與家庭照護造成沉重負擔，故及早且準確的篩檢至關重要。MoCA 為臨床常用工具，但在臺灣情境下，MoCA-T 的命名題可能受教育年數、主要語言使用與文化熟悉度影響，且僅收單一標準答案的規則可能低估真實能力。本研究以MoCA-T命名為主要分析項目，檢驗性別、教育、主要語言使用、經濟狀況與認知嚴重度對作答的影響，並加入台灣常見動物作為補充刺激，評估文化熟悉度是否提升表現，以提供臨床判讀與量表本土化之實證依據。研究目的：本研究旨在探討影響認知障礙症患者在蒙特利爾認知評估量表台灣版(MoCA-T)命名題目作答表現的潛在因素。研究方法及資料：研究方法採用質性與量化並行，包括MoCA-T施測、表現觀察、錯誤類型分析、補充動物圖卡測試，並以單因子變異數分析等統計方法檢視不同背景變項對作答表現的影響。研究結果：本研究共納入71位長者（女性49、男性22），平均年紀79.3歲，(1)命名項目整體表現正確率為獅子54.9%、犀牛63.4%、、駱駝 54.9%。(2) 各題答對率之分組比較：三題均呈男性作答正確率＞女性作答正確率；教育程度：整體呈正向關聯，教育程度較高、雙語（國語、台語皆可使用）背景或對測驗動物較熟悉的長者正確率較高；在補充加入台灣在地文化常見動物的測試時，長者的命名表現優於原有MoCA-T題目，凸顯文化熟悉度對命名作答的影響。(3)錯誤類型分析發現長者的常見命名錯誤包含語意提取失誤、視覺辨識困難、以描述代替命名，以及無法作答等。MoCA-T命名題目可能因文化不熟悉而限制長者正確作答。研究限制：本研究樣本來自北部三家日照中心，且以低學歷、台語為主要語言之長者為主，具區域性與語言、教育結構偏倚，限制結果外推至全臺不同族群（如客語、原住民族群）之廣泛性。另因長照服務實際使用樣態所致，樣本呈女性比例偏高、男性較少，可能影響性別比較之穩定性。補充之文化動物圖片未完成標準化，可能會有干擾因素。結論與建議：未來建議(1)擴大語言與族群多樣性：建議採跨區域（都市、郊區、東部）多點收案，平衡性別與教育分布，並納入多語背景（國語、台語、客語、原住民族語、日語等），以提升外部效度與跨族群比較價值。(2)發展本土化量表：建立具在地代表性的動物與生活物件圖片庫，量化其可辨性與詞彙熟悉度並完成標準化，作為 MoCA-T在地化修訂與計分參照，以提升工具之敏感性與文化適切性。關鍵詞: 認知障礙症；蒙特利爾認知評估量表台灣版(MoCA-T)；命名困難；文化適應性

AbstractBackground and Motivation:Neurocognitive disorder is a progressive condition accelerating with population ageing and imposes substantial burdens on health systems and families. Early and accurate screening is therefore critical. The Montreal Cognitive Assessment–Taiwanese version (MoCA-T) is widely used clinically; however, its Naming subtest may be influenced by years of education, primary language use, and cultural familiarity. In addition, a “single correct answer” scoring rule may underestimate true ability. This study focused on the MoCA-T Naming items, examining the effects of sex, education, primary language, socioeconomic status, and disease severity, and introduced Taiwan-familiar animal stimuli to evaluate whether cultural familiarity improves performance—providing evidence to inform clinical interpretation and localization of the instrument.Objective:This study aimed to explore the potential factors that influence the naming performance of individuals with cognitive impairment on the Taiwan version of the Montreal Cognitive Assessment (MoCA-T).Methods:A mixed-methods design combining both qualitative and quantitative approaches was adopted. Data collection included observation of MoCA-T performance, error type analysis, and supplementary culturally relevant animal picture testing. One-way ANOVA and other statistical methods were used to examine the effects of background variables on naming performance.Results: Seventy-one older adults participated (49 women, 22 men; mean age = 79.3 years).(1) Overall Naming performance: item-level accuracy was lion 54.9%, rhinoceros 63.4%, and camel 54.9%.(2) Group comparisons by background: across items, men outperformed women; education showed a positive association with accuracy; participants with bilingual ability (able to use Mandarin and Taiwanese) or with greater familiarity with the target animals performed better. When Taiwan-familiar animals were added, naming performance exceeded that on the original MoCA-T items, highlighting the impact of cultural familiarity on naming.(3) Error types: the most frequent errors were semantic retrieval failures, visual misidentification, descriptive responses in place of naming, and no response. The findings suggest that unfamiliar cultural content in MoCA-T Naming may constrain accurate responses for some participants.Limitations:Participants were recruited from three day-care centers in northern Taiwan and were predominantly lower-educated and Taiwanese-speaking, limiting generalizability to other groups (e.g., Hakka or Indigenous language users). Because of long-term-care utilization patterns, the sample was female-skewed with fewer men, which may affect the stability of sex comparisons. The supplemental culturally familiar animal pictures were not fully standardized, introducing potential confounds.Conclusion: (1) Broaden linguistic and population diversity: conduct multi-site recruitment across urban, suburban, and eastern regions; balance sex and education distributions; and include multiple language backgrounds (Mandarin, Taiwanese, Hakka, Indigenous languages, Japanese) to enhance external validity and cross-group comparisons.(2) Develop localized measures: establish a standardized, Taiwan-relevant picture bank of animals and daily objects, with quantified discriminability and lexical familiarity, to guide MoCA-T localization and scoring and to improve sensitivity and cultural appropriateness.Keywords:Cognitive impairment; Montreal Cognitive Assessment–Taiwan (MoCA-T); naming difficulty; cultural adaptation