護理紀錄稽核量表之建構及其信效度研究

Validity and Reliability of a Nursing Process Scale

背景：具有信效度的護理紀錄稽核工具，可以確保護理紀錄完整，提升病人的照護品質。 目的：建立具有信效度的護理紀錄稽核工具。 方法：本研究利用結構方程模型分析，來自某醫學中心「護理紀錄稽核量表」資料，共660份。依項目生成和內容有效性測試、項目分析與有效性測試、可靠性測試等三階段進行分析。 結果：研究依稽核表之信效度結果與臨床實務整合後，刪除因素負荷量低的項目，並以轉軸後成份矩陣，重新編排各因子內的項目，完成修訂「護理紀錄稽核量表」。在探索性因素分析其Cronbach’s α為.653，Kaiser-Meyer-Olkin檢驗值為.614，巴特利球形檢定（Bartlett’s test of sphericity）結果亦呈現顯著，萃取出五個因子，22題；再經驗證性因素分析模式修正後，縮減為三個因子，10題，所有指標亦達理想標準。因考量臨床的廣度，故最終重要項目由32題改為22題，其中包含驗證性因素分析建議的10題。 結論／實務應用：發展精簡且高信效度之稽核工具，不僅可以節省稽核者時間，亦能有效的評值護理紀錄的品質，提高紀錄的完整性，更重要的是此稽核表通過護理主管會議審查，已於42個臨床病房全面使用。

Background: A valid and reliable nursing record audit tool can simplify nursing records and provide a basis for quality auditing. Purpose: To ensure the validity and reliability of the Nursing Process Scale to promote accurate monitoring of nursing record quality. Methods: This study employed structural equation modeling to examine the content validity and reliability of the current Nursing Process Scale. A total of 660 results from a medical center were used to revise the content and then the validity and reliability of the revised scale were analyzed. Exploratory factor analysis (EFA) and confirmatory factor analysis (CFA) were used in three stages, namely item generation and content validity testing, item analysis and validity testing, and reliability testing. Results: Validity, reliability, and organization based on clinical practice were used to identify and remove scale items with low factor loadings. The remaining items were organized under several factors in the revised Nursing Process Scale, which had good internal consistency with a Cronbach’s α of .653 in the EFA, a Kaiser-Meyer-Olkin value of .614, and a significant Bartlett’s test of sphericity value. Five factors and 22 questions were extracted from the original 32 questions. The CFA conducted after the model correction reduced the number of questions to 10 and the number of factors to 3, with each index reaching the ideal level. To improve ease-of-use in clinical settings, the important items were reduced from 32 to 22 questions, including the 10 questions suggested by the CFA. Conclusion: The validity, reliability, and organization based on clinical practice were considered in the removal of items with low factor loadings. Axial conversion was used to generate a component matrix, which allowed item rearrangement across factors and the revision of the Nursing Process Scale. The development of simple valid and reliable audit tools will save auditor time and allow the effective evaluation of nursing record quality and improvement in record integrity. This revised scale was reviewed and approved for implementation in 42 clinical wards.