簡短式心智功能檢核量表用於中風住院病人之實用功能檢驗

Abstract

研究背景與目的：認知功能損傷為中風病人常見之後遺症，且對於復健住院期間病人後續之功能恢復影響甚鉅。簡短式心智功能檢核量表 (Mini-Mental State Examination, MMSE)為國內外臨床常用之認知評估工具，具備精簡、施測計分簡單、涵蓋層面廣泛等特點。當使用MMSE作為中風病人認知評估工具時，應具備健康相關評估工具所需之區辨、預測及療效評量三項實用功能 (Practical functions)，可以檢驗其分數分布情形、地板�天花板效應、預測效度與反應性等確認之。然而，過去研究並未有足夠證據支持。因此，本研究之目的為檢驗MMSE應用於中風住院復健期病人之區辨、預測及療效評量三種實用功能。 方法：本研究為回溯型研究，資料來自「中風住院病人復健資料庫之建置」計畫。樣本為連續住進台大醫院復健病房之中風病人，樣本篩選標準為年滿18歲，診斷為出血或梗塞型中風；若病人有影響認知的其他重大疾病（如：失智症、精神分裂症等）、無法遵從口語指令、前後測間隔未滿七日等情況，則排除之。參加者於入住復健病房後三天內與出院前三天內各接受一次入、出院評估，並於中風一年後接受一次電訪追蹤，內容包括：MMSE及巴氏量表 (Barthel Index, BI)等。以描述性統計分析病人之人口學特性及MMSE分數分布情形、地板�天花板效應，以檢驗MMSE是否具區辨功能；以BI為外在效標檢驗MMSE之預測效度及外在反應性，以效應值、標準化反應平均值等指標檢驗MMSE之內在反應性。 結果：共168位病人符合本研究篩選條件且完成入出院兩次評估，其中124位完成中風一年後的電訪追蹤。結果顯示，個案於出院時之MMSE得分分布較入院時不均，且具有明顯天花板效應。MMSE之預測效度方面，以個案出院時之MMSE分數預測中風一年後的BI分數具備良好預測效度 (Spearman’s ρ=0.57)，以入院MMSE分數預測出院及中風一年後之BI分數也具有可接受的預測效度 (ρ=0.40, 0.45)。MMSE用於中風住院復健病人具備小至中度的內在反應性 (ES=0.33; SRM=0.66)；外在反應性則為小效應（以BI為效標，ρ=0.26）。 討論與結論：本研究首次以中風住院病人為對象，檢驗MMSE之區辨、預測及療效評量三種實用功能。研究結果顯示MMSE用於中風病人入院時區辨功能較用於出院時佳；相對於入院MMSE分數，以病人出院之MMSE分數預測病人中風一年後BADL獨立程度之預測功能較佳；療效評量方面，MMSE用於中風住院病人具備低度至中度內在反應性，可適當反應病人住院期間認知功能變化。本研究支持MMSE可作為臨床人員於個案復健期間之治療計畫參考及療效評量指標，而個案出院時的MMSE分數則可做為擬定病人出院及居家計畫之參考。

Background and purposes: Cognitive impairments, one of the common sequelae in patients with stroke, often have huge impact on patients’ functional recovery in multiple domains. Mini-Mental State Examination (MMSE) is a widely-used and simple evaluation tool for assessing patients’ general cognitive function. Thus far, however, in the context of cognitive evaluation for stroke patients, MMSE has not been examined for its score distribution, floor and ceiling effects, predictive validity, and responsiveness, which limits its practicability. The current study therefore aims to examine the score distribution, floor and ceiling effects, predictive validity, and responsiveness of MMSE in inpatients with stroke to validate its discriminative, predictive and evaluative functions. Methods: We took a retrospective study approach in the current research. All data were acquired from the database of “Establishment of a databank for inpatients receiving stroke rehabilitation”. Eligible participants were recruited from consecutive admissions with stroke to the rehabilitation wards at National Taiwan University Hospital. Participants included adults aged over 18 y/o with an ischemic or hemorrhagic stroke. We excluded patients who has history of neurological or psychiatric diseases, such as dementia, schizophrenia, unable to follow commands, and who stayed in the rehabilitation wards for less than 7 days. Both MMSE and Barthel Index (BI) were administered to each participant within 3 days after admission, before discharge from the rehabilitation ward, and one year after stroke onset. We examined the score distribution, ceiling and floor effects of MMSE. BI was used as an external criterion for examining the predictive validity, and external responsiveness of MMSE. Moreover, the responsiveness of MMSE was measured by the effect size (ES) and the Standardized response mean (SRM). Results: Our data analysis included 168 patients, and 124 of them completed the 1-year-after-onset follow-up. The results showed that the MMSE displayed a notable ceiling effect at discharge, while the ceiling effect was not observed at admission. The predictive validity of MMSE was between acceptable to good (Spearman’s ρ=0.40-0.57). Using BI score as external criterion, the external validity of the MMSE was poor (ρ=0.26). The changes in scores of MMSE between admission and discharge were between small to medium (ES=0.33; SRM=0.66), indicating that the MMSE had small to medium internal responsiveness. Discussion and Conclusion: This study is the first to examine the three practical functions of MMSE in the clinical context with stroke inpatient. Our findings suggest that MMSE has sufficient discriminability while administrated at admission, good predictive function while using the MMSE score at discharge to predict the BADL function 1-year after the patients’ onset. Furthermore, MMSE has small to medium internal responsiveness, indicating that MMSE is appropriate for prognosis evaluation. The evidence supports that MMSE is a proper assessment tool with discriminability, predictability and evaluative function for assessing cognitive function of stroke inpatients.