巴氏量表與改良式巴氏量表應用於中風病人之反應性比較

背景：日常生活功能為中風病人常見之失能，進而影響生活品質，因此為臨床介入之主要指標，能協助臨床人員擬定介入計畫及後續照護需求之決策。臨床常用之日常生活功能評估工具，如巴氏量表 (Barthel Index, BI) 及改良版巴氏量表 (Modified Barthel Index, MBI)，能區辨中風病人之功能程度，協助擬定治療計畫，利於臨床人員作為療效評估之指標。評估工具的反應性為瞭解介入成效必備之心理計量特性，而MBI之反應性尚欠缺實證，影響臨床及研究人員對其評量結果之解釋。因此本研究目的為比較BI與MBI之反應性，作為臨床或研究選擇之參考。方法：招募44位中風病人，受試者將於住院一週內進行第一次評估，間隔三週後（21-28天）進行第二次評估，再間隔三週後進行第三次評估。共計算三個評估區間之反應性，分別為：前期（第一次至第二次評估）、後期（第二次至第三次評估）及全期（第一次至第三次評估）。團體層級反應性以凱濟斯效應值 (Kazis’ effect size) 及標準化反應平均值 (standardized response mean, SRM) 進行比較，並以配對t檢定 (paired t-test) 檢驗；個別層級反應性指標為超過最小可偵測變化值 (minimal detectable, MDC) 之人數比例。結果：共44、44與33位病人完成第一次、第二次及三次評估。在團體層級反應性部分，BI三評估區間（前期、後期及全期）之效應值介於0.24-0.65，SRM介於0.69-1.30；MBI效應值介於0.23-0.61，SRM介於0.97-1.63。個別層級反應性部分，改變量超過MDC95者，三區間BI分別為20.5%、15.2、60.6%；MBI則為31.8%、12.1%、60.6%。結論：BI與MBI應用於急性中風病人之個別層級反應性相似，MBI之團體層級反應性略優於BI，本研究建議宜使用MBI以充分呈現病人之功能改變。本研究結果可提供臨床及研究人員合理解讀病人前後測分數的變化，以及選擇適當評估工具。

關鍵字

巴氏量表；改良式巴氏量表；中風；反應性

並列摘要

Introduction: Dysfunction in Activities of daily living (ADL) is a common disability for patients with stroke, and it is related to quality of life. Assessing ADL functioning is important for clinicians in intervention planning and decision making of healthcare needs. The Barthel Index (BI) and Modified Barthel Index (MBI) are commonly used in clinical environments. They can be outcome indicators to provide information on the level of independence. However, the group- and individual-level responsiveness of the MBI remains unclear. Thus, the purpose of this study was to compare the group- and individual-level responsiveness of the MBI to those of the BI. Method: A total of 44 participants were recruited. The participants were administered with the BI and MBI three times with 3 week intervals. The first assessment was administered within one week after admission. We calculated the responsiveness of the BI and MBI in interval 1 (from assessment 1 to assessment 2), interval 2 (from assessment 2 to assessment 3) and the overall period (from assessment 1 to assessment 3). Group-level responsiveness was examined using Kazis’ effect size (ES), standardized response mean (SRM), and paired t-test. Individual-level responsiveness was examined through the proportion of people whose change in ADL function was larger than the minimal detectable change (MDC). Result: There were 44, 44 and 33 participants who completed the BI and MBI at assessment 1, 2, and 3, respectively. In terms of group level responsiveness, the ES of the BI (in interval 1, interval 2, and overall period) was 0.24-0.65, while the SRM was 0.69-1.30. The ES of the MBI was 0.23-0.61, while the SRM was 0.97-1.63. For individual-level responsiveness, the number of participants administered with the BI exceeding the MDC was similar to that of those administered with the MBI in the overall period. Conclusion: MBI had better group-level responsiveness than the BI while the individual-level responsiveness was similar. In sum, we suggest using the MBI for detecting the changes in ADL functioning for patients with stroke.