背景和目的：本研究旨在探討中風病患出院返家後三個月內之跌倒立即機轉，並找出此關鍵轉銜
期之跌倒高危險群預測因子。方法：十九位中風病患在出院前ㄧ週內接受完整的認知與身體檢查
評估，以作為可能的跌倒高危險群預測因子。出院後三個月內接受每兩週一次的跌倒情形追蹤，
若有跌倒發生則安排家訪，以分析跌倒發生之情境以及生物力學相關之立即機轉（身體質心或支撐底面積之干擾）。結果：二十五次跌倒事件中，執行動態功能性活動時之身體質心與支撐底面積之干擾分別佔了76%與20%的立即機轉，並且大部分的跌倒事件（76%）發生在熟悉環境中執行規律活動時。出院前富格梅爾量表之上、下肢動作功能分數分別低於46與28分，最能有效預測出院後跌倒高危險群（敏感度≥80%，特異度≥88%）。結論：中風病患出院後之跌倒主要歸因於執行日常規律活動時身體質心受干擾。因此，出院前之治療計畫應針對富格梅爾量表動作功能分數表現差之高危險群病患，評估與訓練其身體質心控制能力。

Background and Purpose: This preliminary study investigated the immediate mechanisms leading to falls in the first three months after patients with stroke were discharged from hospital and identified the predictors of fallers. Methods: Post-discharge falls of nineteen patients with stroke were monitored in biweekly follow-ups up to three months after discharge, and home visits were arranged whenever falls occurred. The immediate mechanism of each fall was categorized according to the body part primarily perturbed during the fall, i.e. the centre-ofmass (COM) or base-of-support (BOS). Cognitive and physical performances were assessed in
the week before discharge, serving as potential predictors of fallers. Results: Seventy-six and twenty percent of all 25 falls were due to COM and BOS perturbations, respectively. Seventysix percent of the falls occurred during routine activities in familiar environments. Poorer motor
function (upper extremity ≤ 46/66, lower extremity ≤ 28/34 of the Fugl-Meyer Assessment)proved to be the best overall predictor (sensitivity ≥ 80%, specificity ≥ 88%) of individuals prone to falls. Conclusions: The majority of the falls of recently discharged patients with stroke
were due to COM perturbations while performing daily functional activities. Pre-discharge programmes should target at improving COM control in high-risk patients of post-discharge falls, i.e. those with poorer motor function as revealed by the Fugl-Meyer Assessment.