目的:比較原版柏格氏平衡量表 (Berg Balance Scale, BBS) 與短版柏格氏平衡量表(Short-Form Berg Balance Scale, SFBBS) 應用於中風病人在團體層級與個別層級之反應 性。方法:利用次級資料分析。樣本為亞急性中風病人，且於中風發病後第 14、30 及 90天各接受一次 BBS 之評量。團體層級反應性以 BBS 及 SFBBS 之「效應值」(effect size, ES) 及「標準化反應平均值」(standard response mean, SRM) 進行比較;個別層級反應 性則以「最小可偵測變化值」(minimal detectable change, MDC) 為切點，當病人的變化 超過「最小可偵測變化值」則視為具有統計顯著性(真實的改變)。比較二量表分數超 過「最小可偵測變化值」之人數以及「最小可偵測變化值比值」(改變量/最小可偵測 變化值)之差異。結果:共 226、202 與 168 位中風病人完成第 14 天、第 30 天及第 90天之評估。SFBBS 的分數來自病人之 BBS 分數。團體層級反應性檢定結果顯示 BBS 與SFBBS 之「效應值」(介於 0.37 至 0.77)與「標準化反應平均值」(介於 0.64 至 1.05) 相當。個別層級反應性在各不同受測區間，SFBBS 之「最小可偵測變化值比值」均顯著 大於 BBS。結論:在團體層級，原版 BBS 與 SFBBS 之反應性相當。個別層級之反應性，SFBBS 之「最小可偵測變化值比值」顯著大於原版 BBS。基於本研究結果，臨床上可優 先考慮使用 SFBBS。後續研究可以針對此二評估工具之個別層級反應性以不同樣本或不 同方法進行驗證。

Objective: To compare the group-level and individual-level responsiveness of the original Berg Balance Scale (BBS) and the Short-Form Berg Balance Scale (SFBBS) in stroke patients. Method: Analysis of a secondary data set. The participants were stroke patients on subacute stage. The BBS was administrated on participants at the fourteenth, thirtieth, and ninetieth days after stroke onset. The “effect size” (ES) and “standard response mean” (SRM) were used to compare the group-level responsiveness of the BBS and the SFBBS. The individual-level responsiveness was calculated based on the value of minimal detectablechange (MDC).If the patient’s changed score was greater than the MDC of the BBS or SFBBS, the patient’s change was considered significant. The difference in the number of patientsscoring greater than the MDC and the units of MDC (MDC ratio) improved by the patients on both measures was examined. Results: A total of 226, 202, and 168 stroke patients completed the assessments at 14, 30, and 90 days after stroke, respectively. The scores of the SFBBS were extracted from the scores of the BBS of the patients. At the group-level, the BBS and the SFBBS had similar responsiveness, with ES ranging between 0.37 and 0.77 and SRM ranging between 0.64 and 1.05. At individual level, the MDC ratio of the SFBBS was significantly larger than that of the BBS at each observed interval of 14-30 days, 14-90 days, and 30-90 days. Conclusion: At the group-level, the BBS and the SFBBS had similar responsiveness.At the individual-level, the MDC ratio of the SFBBS was signi cantly greater than that of theBBS. Thus, according to our result, we recommend that the SFBBS would be used in clinical settings. Future researches are warranted to cross validated the responsiveness of these two measures with various samples and methods.