術後及創傷後譫妄的發生率高達74-75%，其譫妄病程短且症狀波動，若未經確診或被誤診，可能會引起其他醫療問題。現有評估工具CAM-ICU與ICDSC難以及時診斷及預防。本創傷加護病房單位自2018年起，使用ICDSC量表評估，然而，在2018至2021年間，譫妄發生率仍維持在57-69%，這激發團隊成員積極尋找如何及早監測、及早預防譫妄照護的方法。藉由實證醫學（evidence based medicine, EBM）的5大步驟：提出聚焦問題、搜尋科學證據、嚴格評讀、將加護病房譫妄預測模型（PRE-DELIRIC），及譫妄組合式照護措施，進行知識轉譯，我們提出了「Sweet SMART home」的多元組合式譫妄照護，應用於病人身上，並對過程進行成效評值。PRE-DELIRIC的早期預警，改變了以往的譫妄評估與診斷，於此基礎上，我們針對不同風險值的病人，導入相對應的「Sweet SMART home」多元組合式譫妄照護，進一步有效地降低了譫妄的發生率。這一系列措施成功地將譫妄發生率，由57.8%降至21.5%。

The study indicates that the incidence of delirium related to post-operative and trauma is as high as 74-75%. Delirium is characterized by its short course and the symptoms of fluctuation. It is possible to lead to some other medical issues when the delirium is not firmly diagnosed or misdiagnosed. Current assessment tools-CAM-ICU and ICDSC- can hardly provide timely diagnosis and prevention. The unit of trauma ICU has been using the ICDSC scale to assess delirium in patients since 2018. However, the delirium incidence in the unit ranged from 57% to 69% between 2018 and 2021. Such a condition stimulated the team members to engage in evidence-based activities to find ways to monitor and prevent delirium early. We translated the ICU delirium prediction model (PRE-DELIRIC) and a combination of care measures into knowledge. Then, we applied them to our trauma surgical ICU, submitting the "Sweet SMART home" multifaceted delirium assessment and care approach through the five steps of EBM (Evidence-Based Medicine): Generate clinical questions, find the best evidence, critically appraise, apply the evidence, and evaluate information, skill, and EBM procedures. The early warning provided by PRE-DELIRIC changed conventional delirium assessment and impromptu diagnosis. The corresponding "Sweet SMART home" multifaceted delirium assessment and care were introduced to patients at various risk levels, effectively reducing the incidence of delirium. A series of actions successfully reduced the incidence of delirium from 57.8% down to 21.5%.