Chen Hong-Lin, Cao Ying-Juan, Zhang Wei, Wang Jing, Huai Bao-Sha
School of Nursing, Nantong University, Nantong City, Jiangsu Province, PR China.
Qilu Hospital of Shandong University, Jinan City, Shandong Province, PR China.
Appl Nurs Res. 2017 Feb;33:169-174. doi: 10.1016/j.apnr.2016.12.001. Epub 2016 Dec 5.
The inter-rater reliability of Braden Scale is not so good. We modified the Braden(ALB) scale by defining nutrition subscale based on serum albumin, then assessed it's the validity and reliability in hospital patients.
We designed a retrospective study for validity analysis, and a prospective study for reliability analysis. Receiver operating curve (ROC) and area under the curve (AUC) were used to evaluate the predictive validity. Intra-class correlation coefficient (ICC) was used to investigate the inter-rater reliability.
Two thousand five hundred twenty-five patients were included for validity analysis, 76 patients (3.0%) developed pressure ulcer. Positive correlation was found between serum albumin and nutrition score in Braden scale (Spearman's coefficient 0.2203, P<0.0001). The AUCs for Braden scale and Braden(ALB) scale predicting pressure ulcer risk were 0.813 (95% CI 0.797-0.828; P<0.0001), and 0.859 (95% CI 0.845-0.872; P<0.0001), respectively. The Braden(ALB) scale was even more valid than the Braden scale (z=1.860, P=0.0628). In different age subgroups, the Braden(ALB) scale seems also more valid than the original Braden scale, but no statistically significant differences were found (P>0.05). The inter-rater reliability study showed the ICC-value for nutrition increased 45.9%, and increased 4.3% for total score.
The Braden(ALB) scale has similar validity compared with the original Braden scale for in hospital patients. However, the inter-rater reliability was significantly increased.
Braden量表的评分者间信度不太理想。我们通过基于血清白蛋白定义营养子量表对Braden（ALB）量表进行了修改，然后评估其在住院患者中的有效性和信度。
我们设计了一项回顾性研究用于有效性分析，以及一项前瞻性研究用于信度分析。采用受试者工作特征曲线（ROC）和曲线下面积（AUC）来评估预测有效性。组内相关系数（ICC）用于研究评分者间信度。
纳入2525例患者进行有效性分析，76例患者（3.0%）发生了压疮。Braden量表中血清白蛋白与营养评分呈正相关（Spearman系数0.2203，P<0.0001）。Braden量表和Braden（ALB）量表预测压疮风险的AUC分别为0.813（95%CI 0.797 - 0.828；P<0.0001）和0.859（95%CI 0.845 - 0.872；P<0.0001）。Braden（ALB）量表比Braden量表更有效（z = 1.860，P = 0.0628）。在不同年龄亚组中，Braden（ALB）量表似乎也比原始Braden量表更有效，但未发现统计学显著差异（P>0.05）。评分者间信度研究表明，营养方面的ICC值增加了45.9%，总分增加了4.3%。
对于住院患者，Braden（ALB）量表与原始Braden量表具有相似的有效性。然而，评分者间信度显著提高。
