Liu Fang, Tsang Raymond Cc, Zhou Jing, Zhou Mingchao, Zha Fubing, Long Jianjun, Wang Yulong
Department of Rehabilitation, Shenzhen Second People's Hospital/Health Science Centre, The First Affiliated Hospital of Shenzhen University, Shenzhen, People's Republic of China.
Department of Physiotherapy, MacLehose Medical Rehabilitation Centre, Hong Kong Special Administrative Region, People's Republic of China.
J Stroke Cerebrovasc Dis. 2020 Sep;29(9):105033. doi: 10.1016/j.jstrokecerebrovasdis.2020.105033. Epub 2020 Jun 23.
Modified Rankin Scale and Barthel Index are the most common scales for assessing stroke outcomes in clinical practice and trials. Concordance between the Barthel Index scores and the modified Rankin Scale grades is important to define favorable outcome in clinical practice and stroke trials consistently. The purpose of this study was to examine the relationship between the scores of Barthel Index and 3-item Barthel Index Short Form with the modified Rankin Scale grades of acute stroke patients.
Barthel Index, Barthel Index Short Form scores and modified Rankin Scale grades of 516 stroke patients were obtained from a follow-up study of the Longshi Scale in China. A study showed that the assignment of modified Rankin Scale grades to stroke patients was prone to misclassification. Therefore, the recorded modified Rankin Scale grades were compared with the Barthel Index scores of each patient to produce the adjusted modified Rankin Scale grades. Receiver operating characteristics curve analyses were performed to determine the optimal cutoff scores, respective sensitivities and specificities of Barthel Index and Barthel Index Short Form with the corresponding adjusted modified Rankin Scale grades ≤1, ≤2 and ≤3.
About 44% of the recorded modified Rankin Scale grades of patients required adjustment. The optimal cutoff scores were ≥100 (sensitivity 100%; specificity 95.3%), ≥100 (sensitivity 98.1%; specificity 100%) and ≥75 (sensitivity 93.8%; specificity, 91.9%) for the Barthel Index and ≥40 (sensitivity 100%; specificity 78.9%), ≥40 (sensitivity 98.1%; specificity 82.8%), and ≥35 (sensitivity 99.3%; specificity, 91.6%) for the Barthel Index Short Form corresponding to the adjusted modified Rankin Scale grade ≤1, ≤2, and ≤3 respectively. The areas under the receiver operating characteristic curves were nearly all above 0.9.
The optimal cutoff scores of Barthel Index and Barthel Index Short Form corresponding to the modified Rankin Scale grades ≤1, ≤2 and ≤3 were recommended to be ≥100 and ≥40, ≥100 and ≥40, and ≥75 and ≥35 respectively for determining the favorable and unfavorable outcome of stroke patients within three months of onset in clinical practice and trials.
改良Rankin量表和Barthel指数是临床实践和试验中评估卒中预后最常用的量表。Barthel指数评分与改良Rankin量表分级之间的一致性对于在临床实践和卒中试验中始终如一地定义良好预后很重要。本研究的目的是探讨急性卒中患者的Barthel指数评分及3项简化Barthel指数与改良Rankin量表分级之间的关系。
从中国龙氏量表的一项随访研究中获取了516例卒中患者的Barthel指数、简化Barthel指数评分及改良Rankin量表分级。一项研究表明,给卒中患者分配改良Rankin量表分级容易出现错误分类。因此,将记录的改良Rankin量表分级与每位患者的Barthel指数评分进行比较,以得出调整后的改良Rankin量表分级。进行受试者工作特征曲线分析,以确定Barthel指数和简化Barthel指数在相应调整后的改良Rankin量表分级≤1、≤2和≤3时的最佳截断分数、各自的敏感性和特异性。
约44%患者记录的改良Rankin量表分级需要调整。对于Barthel指数,对应调整后的改良Rankin量表分级≤1、≤2和≤3时,最佳截断分数分别为≥100(敏感性100%;特异性95.3%)、≥100(敏感性98.1%;特异性100%)和≥75(敏感性93.8%;特异性91.9%);对于简化Barthel指数,最佳截断分数分别为≥40(敏感性100%;特异性78.9%)、≥40(敏感性98.1%;特异性82.8%)和≥35(敏感性99.3%;特异性91.6%)。受试者工作特征曲线下面积几乎均大于0.9。
在临床实践和试验中,对于确定发病三个月内卒中患者的预后好坏,推荐改良Rankin量表分级≤1、≤2和≤3时,Barthel指数及简化Barthel指数的最佳截断分数分别为≥100和≥40、≥100和≥40、≥75和≥35。