全靜脈營養為腸胃道功能障礙或無法接受腸道營養病人重要的支持性療法，但卻可能誘發高血糖、 感染等不良生理反應。許多研究證明嚴格控制重症病人的血糖可以降低感染率、加護病房住院天數與死亡 率。而本院於2006年8月依據美國糖尿病學會之血糖控制指引將全靜脈營養病人的血糖控制範圍，由原 本180 mg/dL-250 mg/dL降至110-180 mg/dL ’然而’對於接受全靜脈營養治療容易引發高血糖的病人’嚴 格血糖控制與住院期間的併發症之相關研究並不多見。因此，本研究目的在於探討嚴格血糖控制對接受全 靜脈營養治療病人併發症之影響，採非同期比較性研究設計，研究對象以病歷回溯方式收集本院2004年6 月至2008年7月接受全靜脈營養治療病人共323人，排除不符合篩選條件者後，共納入研究者189人， 評估不同血糖控制標準(180 mg/dL-250 mg/dL的病人97人，與110-180 mg/dL的病人92人），對降低全靜 脈營養治療病人臨床併發症之成效。結果顯示在全靜脈營養治療期間，兩組採不同血糖控制標準的病人發 生急性腎衰竭、感染與敗血症的比率皆未達統計顯著差異(p>0.05)。多元邏輯斯迴歸分析顯示急性生理及 慢性健康評估分數(Acute Physiology and Chronic Health Evaluation II scores, APACHE II scores) (0R=1.05, 95% CI=1.00-1.10)、全靜脈營養治療天數(OR=1.02，95%CI=1.00-1.04)和治療期間平均血糖高於180 mg/dL (OR=3.10, 95%CI=1.08-8.89)都是導致病患死亡的危險因子。但是血糖控制範圍110-180 mg/dL組的低血糖 發生率卻高於180-250 mg/dL組(46.7% vs. 7.2%, p<0.001)，低血糖發生與感染之風險有關，唯本研究無法 證實低血糖的發生與本組病人死亡率之相關性。本研究之結論可初步作為預防全靜脈營養治療對病人不良 預後之參考，未來應設計前瞻性研究來確認其進一步之結論。

Total parenteral nutrition (TPN) is a major supportive treatment in critically ill patients who could not tolerate enteric feeding. TPN treatment had been associated with adverse effects including infection and hyperglycemia. Numerous studies have documented that strict glycemic control in critical ill patient can reduce infection, length of stay in intensive care units and mortality. However, there are fewer reports regarding the associations of hyperglycemia with strict glycemic control and adverse outcomes during hospitalization in patients receiving TPN. In August 2006, our hospital followed the American Diabetes Association guidelines adjusting range of glycemic control from 180-250 mg/dL to 110-180mg/dL in patients treated with TPN. The purpose of the present study was to examine the effects of strict glycemic control on complications in patients receiving TPN treatment. In this non-concurrent comparison study we retrospectively reviewed the chart records of 323 patients who received TPN treatment during the period between June 2004 and July 2008. A total of 189 patients were recruited after excluding those who did not match the inclusion criteria. The effectiveness of different extent of strict blood glucose control (target of 180-250 mg/dL before August 2006, n=97 and target of 110-180mg/dL after August 2006, n=92) in reducing the clinical complications during TPN use were examined. Our results showed that there was no significant difference in mortality, infections, and sepsis between the two groups (^>0.05) during their hospitalization. In multiple logistic regression models adjusted for age, gender, and diabetes, mortality was predicted by APACHE II scores (0R=1.05, 95%CI=1.00-1.10), duration of TPN use (0R=1.02, 95%CI=1.00-1.04), and mean blood glucose>180 mg/dL during TPN use (0R=3.10, 95%CI=1.08-8.89). Risk of hypoglycemia was higher in the group targeted for 110-180 mg/dL(p<0.001). Hypoglycemia was associated with increased risk of infections (OR=3.13, 95%CI=1.05-9.36), however, hypoglycemia was not found to be related to mortality. Further studies should be carried out prospectively to confirm these findings.