目的：探討護理之家住民長期使用管灌飲食與一般經口進食者間，其營養狀態是否差 異及其相關影響因子。方法：本研究資料來源為病例回塑性病歷回溯，由西元2011年1月1 曰到2011年12月01曰間，居住於北部某醫學中心附設護理之家中，願接受當年度成人預防 保健檢查之住民為收案對象。除了收集健康檢查資料中各項生理、生化數據及特定營養指 標外。同時，還包括住民基本資料、過去慢性疾病史(如：糖尿病、高血壓、高血脂、失智 (包含阿茲海默症及巴金森氏症診斷)及腦血管疾病（包含腦中風及腦出血）、壓瘡及憂鬱症等 七項常見慢性疾病）、定期迷你營養評估篩檢量表(Mini nutritional assessment-short form, MNA-SF)、以及於體檢後三個月内急診次數；統計方法包括描述性分析、魏可遜二様本檢 定(Wilcoxon two-sample test)、卡方同質性檢定及線性迴歸分析。結果：總共收集到62筆住 民資料，其中男性住民共28位(45.2%)，使用管灌飲食者為34位(54.8%)，經迷你營養評估 篩檢量表篩選營養不良者，有51位(82.3%)。比較管灌與經口飲食組住民二者，發現管灌飲 食組在多項營養指標及三個月内急診次數等表現較差，如：熱量需求達成率(TER%)管灌飲 食組為100.32土 10.51%及經口飲食組為109.61 土 11.53%(p<0.000)；小腿圍分別為27.3土 3.2cm，30.1±4.14cm (p=0.013);白蛋白濃度分別為3.52±0.40 g/dl，3.82±0.41g/dl (p=0.010)， 以及白血球數目分別為7.12±1.82 x103/uL，6.00±1.82 x103/uL (p=0.016)，具有統計學上顯 著意義。且使用管灌飲食之住民在接受體檢後三個月内，急診就醫次數遠高於一般經口進 食之住民(0.5±0.9，0.1±0.3，p=0.013)。於三個月内急診次數線性迴歸分析中，則發現C-反 應蛋白為主要影響因子，每增加1 mg/dl，三個月内急診次數則將會增加0.405次(p=0.001)。 結論：本研究之結果發現，長期使用管灌飲食之住民相較經口進食者，整體營養狀態較 差，平均急診就醫次數增加，其原因可能與慢性發炎狀態相關。而各項營養指標中，尤以 C-反應蛋白(C-reactive protein ，CRP)，將有助於醫療人員預測住民急診就醫次數。

Objective: To investigate the risk factors and association of the nutritional status between long-term tube-feeding and oral feeding of institutionalized residents. Methods: This was a prospective analysis of nutritional status from institutionalized residents from January 1st, 2011 to
st
December 01 , 2011. Residents who had received the annual health examination were included in the analysis. We also collected specific nutritional markers and other information of residents, including general demographic data, past medical history (eg. diabetes, hypertension, hyperlipidemia, dementia, cerebrovascular disease, pressure sore and depression), regular mini-nutritional assessment soft-form score (MNA-SF) and the number of emergency visits within 3 months after indexed date. Descriptive analysis, Wilcoxon two-sample test, Chi-square test, and linear regression were performed for statistics. Results: There were 62 institutionalized residents with the annual health examination during collection period. The male accounted 28 (45.2%); tube-feeding subjects accounted 34 (54.8 %); poor nutrition subjects (according to MNA-SF) accounted 51 (82.3%). Notably, the nutritional status of tube-feeding group was more worse than oral-feeding group, such as total energy requirement (100.32±10.51%，109.61±11.53%， p<0.000), parameters of calf circumference(27.3±3.2cm ，30.1±4.14cm ，p=0.013), albumin level(3.52±0.40 g/dl ’ 3.82±0.41 g/dl ’ p=0.010), white blood cell count (7.12±1.82 x103/uL ’
3
6.00±1.82x10/uL，p=0.001) and number of emergency visits after 3 months later(0.5±0.9，0.1±0.3，p=0.013) had significant difference. In linear regression, we also noted the C-reactive protein level (S: 0.507, p=0.001)was the main risk factor for number of emergency visits after 3 months later. Conclusion: The poor nutritional status of tube-feeding institutionalized residents may be associated with chronic inflammation. We could use regular healthy examinational tools for early screening of the institutionalized residents’ nutritional status, especially C-reactive protein level, for early predicting number of emergency visits after 3 months later.