探討內科加護病房譫妄評估正確性與譫妄教育介入之成效

Abstract

研究背景與目的：加護病房譫妄發生率高，使用經驗證之譫妄評估工具(如加護病房譫妄檢查表 (Intensive Care Delirium Screening Checklist, ICDSC))評估重症患者是早期發現及治療的重要關鍵，目前對於ICDSC譫妄評估正確性的研究並不多，故本研究旨於探討加護病房使用ICDSC評估譫妄之正確性及影響因子，此外，進一步檢驗譫妄教育介入對於ICDSC譫妄評估正確性之成效。研究方法：本研究採前後測研究設計(pre-test post-test comparative study)，於2023年9月至11月期間，在北部某醫學中心的三個內科加護病房隨機抽查護理師使用加護病房譫妄篩檢表(ICDSC) 評估的正確性。由接受過譫妄評估訓練的護理師（即黃金標準）在加護病房進行常規篩檢的一小時內獨立使用ICDSC進行譫妄評估，並與加護病房護理人員使用ICDSC的結果進行比較，以評估護理人員使用ICDSC 評估譫妄的正確性，並以Cohen's kappa係數0.8為標準，若未達到標準，則會於2023年12月至2024年2月進行為期三個月的譫妄教育介入，包含個案討論與困難場景回顧、譫妄認知教育以及一對一即時回饋，並於2024年3月至5月期間再次進行後測評估。研究結果：前測共分析了420組配對資料，ICDSC結果weighted kappa係數為0.69，顯示黃金標準與加護病房護理師在ICDSC上的一致性達中等一致性，在ICDSC的八項指標中，除了幻覺與不恰當言語與情緒達高度一致性外，其餘項目一致性皆低於整體kappa值。經過為期三個月的教育介入後，後測共分析了427組配對資料，ICDSC結果weighted kappa提升至0.74，而在ICDSC的八項子項中只有精神運動性興奮或遲鈍(kappa=0.35)與睡眠-清醒週期紊亂(kappa=0.63)未達到高度一致性(kappa>0.8)。此外，除了教育前已屬高度一致性的幻覺與不恰當言語與情緒外，教育介入後，包含ICDSC結果(p=0.019)以及其他子之正確性百分比都有顯著改善(P<0.05)。年齡是否大於65歲(p=0.004)、是否使用約束(p=0.002)以及不同加護病房單位(p<0.001)與ICDSC譫妄評估正確性顯著相關。結論：內科加護病房的ICDSC譫妄評估正確性仍有進步空間，譫妄教育介入可有效提升ICDSC譫妄評估正確性，年齡是否大於65歲、是否使用約束以及不同加護病房單位是影響ICDSC譫妄評估正確性的相關因子。

Background & Objectives: The incidence of delirium in intensive care units (ICUs) is high. Utilizing validated delirium assessment tools (e.g., the Intensive Care Delirium Screening Checklist, ICDSC) is crucial for the early detection and management of delirium. However, research on the accuracy of ICDSC delirium assessments is limited. This study aims to investigate the accuracy and influencing factors of using the ICDSC for delirium assessment in ICUs and to examine the effects of delirium education interventions in improving the accuracy of ICDSC assessments. Method: This study is a pre-test and post-test comparative design. From September to November 2023, random audits were conducted to evaluate the accuracy of the ICDSC assessment in three medical ICUs at a tertiary care hospital in Taipei, Taiwan. A trained nurse specialist, serving as the gold standard of ICDSC, independently performed ICDSC assessments within one hour of the routine screenings conducted by ICU nurses. The results were compared, with a Cohen’s kappa coefficient of 0.8 set as the benchmark for acceptable agreement. When the accuracy of ICDSC assessments failed to meet this benchmark, a three-month educational intervention was implemented from December 2023 to February 2024. The intervention included structured activities such as case discussions, scenario-based feedback, delirium education, and one-on-one real-time feedback to address gaps in assessment accuracy. Following the intervention, post-test evaluations of ICDSC assessment accuracy were conducted between March and May 2024. Results: The pre-test analysis included 420 paired data sets, with the ICDSC results yielding a weighted kappa coefficient of 0.69. This indicates moderate agreement between the gold standard and ICU nurses in ICDSC assessments. Among the eight ICDSC items, only hallucinations and inappropriate mood or speech showed high consistency, while the scores for other items were lower than the overall kappa value. After the three-month educational intervention, the post-test analysis included 427 paired data sets, and the weighted kappa coefficient for ICDSC results improved to 0.74. Among the eight ICDSC subitems, only psychomotor agitation or retardation (kappa=0.35) and sleep-wake cycle disturbances (kappa=0.63) did not achieve high agreement (kappa > 0.8). Furthermore, aside from hallucinations and inappropriate speech or mood, which already exhibited high agreement before the intervention, the accuracy of ICDSC results (p=0.019) and the correctness percentages of other sub-items showed significant improvement (p<0.05) after the intervention. Significant factors associated with the accuracy of ICDSC delirium assessments included age over 65 years (p=0.004), the use of physical restraints (p=0.002), and differences in ICU units (p<0.001). Conclusion: There is still room for improvement in the accuracy of ICDSC delirium assessments in medical ICUs. Delirium education interventions can effectively increase the accuracy of these assessments. Factors influencing the accuracy of ICDSC assessments include age over 65 years, the use of physical restraints, and differences among ICU units.