Kottner Jan, Tannen Antje, Dassen Theo
Charité Universitätsmedizin Berlin Zentrum für Human- und Gesundheitswissenschaften Institut für Medizin-/Pflegepädagogik und Pflegewissenschaft, Berlin, Germany.
Pflege. 2008 Apr;21(2):85-94. doi: 10.1024/1012-5302.21.2.85.
Pressure ulcer risk assessment scales can assist nurses in determining the individual pressure ulcer risk. Although the Braden scale is widely used throughout Germany, its psychometric properties are yet unknown. The aim of the study was to determine the interrater reliability of the Braden scale and to compare the results with those of published data. A literature review was conducted. 20 studies measuring the interrater reliability of the Braden scale were evaluated. Only three of those studies investigated the interrater reliability of single items. The Pearson product-moment correlation coefficient (0.80 to 1.00) was calculated in most studies for an evaluation of the Braden scale as a whole. However, the use of correlation coefficients is inappropriate for measuring the interrater reliability of the Braden scale. Measures of the intraclass correlation coefficient varied from 0.83 to 0.99. The investigation of the interrater reliability of the Braden scale's German version was conducted in a German nursing home in 2006. Nurses independently rated 18 and 32 residents twice. Nurses achieved the highest agreement when rating the items "friction and shear" and "activity" (overall proportion of agreement = 0.67 to 0.84, Cohen's Kappa = 0.57 to 0.73). The lowest agreement was achieved when the item "nutrition" (overall proportion of agreement = 0.47 to 0.51, Cohen's Kappa = 0.28 to 0.30) was rated. For 66% of the rated residents the difference in the obtained Braden scores was equal or less than one point. Intraclass correlation coefficients were 0.91 (95% confidence interval 0.82 to 0.96) and 0.88 (95% confidence interval 0.61 to 0.96). This indicates that the interrater reliability of the Braden scale was high in the examined setting.
压疮风险评估量表可帮助护士确定个体的压疮风险。尽管布拉登量表在德国被广泛使用，但其心理测量特性尚不清楚。本研究的目的是确定布拉登量表的评分者间信度，并将结果与已发表的数据进行比较。进行了一项文献综述。对20项测量布拉登量表评分者间信度的研究进行了评估。其中只有三项研究调查了单个条目的评分者间信度。大多数研究计算了皮尔逊积差相关系数（0.80至1.00），以评估整个布拉登量表。然而，使用相关系数来测量布拉登量表的评分者间信度是不合适的。组内相关系数的测量值从0.83到0.99不等。2006年在一家德国养老院对布拉登量表德文版的评分者间信度进行了调查。护士对18名和32名居民进行了两次独立评分。在对“摩擦力和剪切力”以及“活动能力”条目进行评分时，护士们达成的一致性最高（总体一致比例=0.67至0.84，科恩卡方=0.57至0.73）。在对“营养”条目进行评分时达成的一致性最低（总体一致比例=0.47至0.51，科恩卡方=0.28至0.30）。对于66%的被评分居民，所获得的布拉登评分差异等于或小于1分。组内相关系数分别为0.91（95%置信区间0.82至0.96）和0.88（95%置信区间0.61至0.96）。这表明在本研究环境中，布拉登量表的评分者间信度较高。
