目的·编制慢性肾脏病（chronic kidney disease，CKD）患者安全用药知识-信念-行为（knowledge，attitude，belief and practice，KABP）量表并检验其信度和效度。方法·基于KABP理论，通过系统的文献回顾和半结构化访谈，由科研小组成员综合讨论设计初始条目。采用德尔菲专家咨询法和预调查改进初始条目形成初始量表。纳入上海交通大学医学院附属第九人民医院CKD患者进行问卷调查，使用临界比值法、同质性检验进行条目分析，分别使用探索性因子分析和逻辑分析检验量表的结构效度和内容效度，使用Cronbach′s α系数、折半信度、重测信度检验量表信度。结果·初始量表由3个维度29个条目组成，共发放213份问卷，收集有效问卷209份。条目分析结果显示条目1、2、3、19均有3项以上指标未达标，因此删除这4个条目后进行效度检验。初始量表KMO（Kaiser-Meyer-Olkin）为0.946，Bartlett′s球形检验χ²值为4 554.451（P=0.000），提示总体的相关矩阵有共同因素存在，适合进行因素分析。使用主成分分析法及直交转轴最大变异法进行因素分析，最终删除条目13；根据各因素内的条目变量特性，提取的3个共同因素分别命名为“安全用药行为”“安全用药知识”“安全用药信念”，3个维度的特征值分别为6.565、6.176、3.863；累积方差贡献率为69.18%。全体一致量表水平内容效度（scale-level-content validity index/universal agreement，S-CVI/UA）和平均量表水平内容效度（S-CVI/average，S-CVI/Ave）分别为0.860和0.987。最终量表由3个维度24个条目组成。信度检验显示总量表的Cronbach′s α系数为0.964，折半信度为0.911，重测信度为0.892。结论·该研究编制的CKD患者安全用药KABP量表具有良好的信度和效度。

Objective ·To develop a knowledge, attitude, belief and practice (KABP) scale for safe medication in patients with chronic kidney disease (CKD), and test its reliability and validity. Methods ·Based on the KABP theory, the initial items were discussed and designed comprehensively by research group members through a systematic literature review and semi-structured interviews. The Delphi method of expert panel consultation and the pre-investigation were conducted to improve the initial scale. Next, the patients with CKD in Shanghai Ninth People's Hospital, Shanghai Jiao Tong University School of Medicine were enrolled to complete the initial scale. The item analysis was conducted by critical ratio and homogeneity test. The construct validity and the content validity of scale were tested by exploratory factor analysis (EFA) and logic analysis, respectively. The reliability of the scale was evaluated by Cronbach's α coefficient, split-half reliability, and test-retest reliability. Results ·The initial scale consisted of 29 items with three dimensions. A total of 213 questionnaires were distributed and 209 valid questionnaires were collected. Items No.1, 2, 3 and 19, which did not meet the criteria of three or more indicators, were deleted after item analysis. The KMO (Kaiser-Meyer-Olkin) was 0.946, and the χ² value of Bartlett's spherical test was 4 554.451 (P=0.000). The results suggested that the overall correlation matrix had common factors, which was suitable for factor analysis. The principal component analysis and the maximum variation method of orthogonal rotation axis were adopted. Item No.13 was deleted finally. According to the variable characteristics contained in each factor, three extracted common factors were named as "Practice of safe medication", "Knowledge about safe medication", and "Belief of safe medication", respectively. The eigenvalues of three dimensions were 6.565, 6.176, and 3.863, respectively, and the cumulative variance contribution rate was 69.18%. The scale-level-content validity index (S-CVI)/universal agreement was 0.860. The S-CVI/average was 0.987. The formal scale was formed of 24 items with three dimensions. The α coefficient of the total scale was 0.964. The split-half reliability coefficient was 0.911 and the test-retest reliability was 0.892 for the scale. Conclusion ·The KABP scale of safe medication in CKD patients has good reliability and validity.