背景與目的：本研究以綜合分析（meta-analysis）方式探討強調行走訓練之任務取向巡迴訓練對中風患者行走能力改善的效果。方法：由兩位評論者搜尋自2000年1月到2008年8月MEDLINE,
PubMed, PEDro, CEPS, CINAHL,及Cochrane等電子資料庫中有關行走之巡迴訓練對中風患者行走功能改善效果之原始著作，並由查詢引用之文獻擴大搜尋範圍。以stroke, cerebrovascular disorders,circuit training, group exercise, exercise training, walking, ambulation, locomotion,及rehabilitation為關鍵字搜尋。文獻納入原則為：⑴隨機控制分組實驗⑵與行走有關之巡迴訓練⑶討論與行走功能有關之結果。納入之文章以PEDro量表對研究設計評分。綜合分析結果以效能估量(effect size)來表示。結果：一共納入6篇文獻，PEDro量表分數由7到9分。結果顯示訓練後與行走有關之變項，如六分鐘行走測驗有明顯改善（weighted mean difference=39.66 m, P= 0.003, 95% CI= 13.45~65.87）；行走速度（weighted mean difference=25.06 cm/s, P=0.08, 95% CI=-3.26~53.39）及計時起走測驗（weighted mean difference=-3.14 s, P=0.12, 95% CI=-7.12~0.84）等變項有改善趨勢但未達統計差異。結論：12到57小時的任務取向巡迴訓練對急性和慢性中風患者行走耐力有改善的效果。（物理治療2009;34(3):167-176）

Background and Purposes: Task-oriented circuit training has been shown to improve balance, endurance, mobility and upper extremity function after stroke. This meta-analysis investigated the effect of task-oriented circuit training with an ambulation component on walking function in
people with stroke. Methods: Articles published between January 2000 and August 2008 were searched from the electronic databases of MEDLINE, PubMed, PEDro, CEPS, CINAHL and Cochrane using the following key words: stroke or cerebral vascular disease, circuit training or group exercise or exercise training or rehabilitation, and walking or ambulation or locomotion. References presented in relevant publications were examined to expand the search. The inclusion criteria were: (1) randomized controlled trials; (2) subjects received circuit training with an ambulation exercise component; (3) outcome measures included at least one of the following measures of walking function: the 6-minutes walk test (6MWT), walking speed, or the timed up and go test (TUGT). Each article was appraised using the PEDro scale by two reviewers. The Review Manager 4.2 software was used to calculate the effect sizes for the outcome measures. Results: Only 6 studies had met our criteria and were included in our meta-analysis. The methodological quality of these articles ranged between 7 and 9 on the PEDro scale. The subjects included both subjects with acute and chronic stroke (36 days to 28 years post stroke). The intensity of circuit training ranged between a total of 6 to 57 hours of exercises. The meta-analysis revealed statistically significant improvement for the 6MWT (weighted mean difference= 39.66 m, P= 0.003, 95% CI= 13.45~65.87), but not for the walking speed (weighted mean difference= 25.06 cm/s, P= 0.08, 95% CI= -3.26~53.39) nor the TUGT (weighted mean difference= -3.14, P= 0.12, 95% CI= -7.12~0.84). Conclusions: Task-oriented circuit training with an ambulation component appears to be most beneficial for improving the walking endurance, as shown with the 6MWT, in people at various stages of stroke. The effective range of treatment duration was between 12 to 57 hours. Further study is needed to identify the most efficient treatment component for this effect, and to investigate the potential reasons causing lack of effect on the walking speed and the TUGT. (FJPT 2009;34(3):167-176)