目標：比較長短期失能居家個案之物理治療服務利用及療效之異同，以供日後急性後期照護與長期照護居家物理治療服務政策推動參考。方法：以2012年台北市與新北市無記名居家物理治療資料進行次級分析。走路、洗澡、穿衣與轉位等任一項身體功能，失能超過6個月者定義為長期失能，其餘為短期失能。以t檢定與卡方檢定比較長短期失能個案特徵、物理治療介入次數及巴氏量表得分改善情形，並以羅吉斯迴歸分析各診斷（或處置）個案之失能期長短和功能改善與否之關聯性。結果：717筆個案資料中，長、短期失能個案比例約為7比3。短期失能個案之疼痛、關節受限與轉介前功能較獨立之比例較高。居家物理治療的平均介入次數為5.4 ± 3.2次，長、短期失能個案的服務利用分別為5.5 ± 3.2次及5.2 ± 3.2次，無統計上的顯著差異。介入後，個案的巴氏量表得分平均改善6.0 ± 13.0分，短期失能個案功能改善顯著優於長期失能個案，巴氏量表分別增加8.9 ± 15.6分與4.8 ± 11.7分（p ＜ 0.01）。與長期失能個案相比，短期失能個案功能改善之勝算比，在中風個案為2.28（p=0.007），在關節置換個案為7.28（p=0.011）。結論：長短期失能個案的居家物理治療介入次數相當，但短期失能個案療效較佳。中風、關節置換的短期失能個案進步情況尤其顯著。

Objectives: This study compared differences in the characteristics, utilization, and outcomes of home-based physical therapy (HPT) between clients with long- and short-term disabilities. Methods: A secondary analysis was conducted on data collected from routine records of HPT in the Taipei metropolitan region in 2012. Clients who had disabilities in walking, dressing, bathing, and transferring for ＞ 6 months were classified as having a long-term disability; the remaining clients were classified as having a short-term disability. The data included socio-demographic data, health functional status, intervention sessions, and Barthel Index (BI) scores. We used t-tests, chisquare tests, and multiple logistic regression models to analyze the differences between long- and short-term disabilities. Results: Delinked data of 717 subjects were included. The ratio of longversus short-term clients with disabilities was approximately 7:3. Significantly more clients with short-term than long-term disabilities reported pain, limited range of motion, and less functional dependence before intervention. The intervention frequency (average, 5.4 ± 3.2 sessions) was similar between groups. After HPT intervention, the average improvement in BI score was 6.0 ± 13.0. The improvements in ADL were significantly greater in the clients with short-term disabilities than the clients with long-term disabilities (8.9 ± 15.6 vs. 4.8 ± 11.7, respectively). The odds ratios between clients with short- and long-term disabilities were 2.28 (p=0.007) for stroke patients and 7.28 (p=0.011) for patients who have undergone arthroplasties. Conclusions: The number of intervention sessions between the clients with long- and short-term disabilities were similar. The effectiveness of HPT intervention was better for clients with short- than long-term disabilities, especially for patients who had undergone arthroplasties or had strokes.