Katzan Irene L, Lapin Brittany
From the Center for Outcomes Research and Evaluation (I.L.K., B.L.) and Cerebrovascular Center (I.L.K.), Neurological Institute, Cleveland Clinic, OH.
Stroke. 2018 Jan;49(1):147-154. doi: 10.1161/STROKEAHA.117.018766.
The International Consortium for Health Outcomes Measurement recently included the 10-item PROMIS GH (Patient-Reported Outcomes Measurement Information System Global Health) scale as part of their recommended Standard Set of Stroke Outcome Measures. Before collection of PROMIS GH is broadly implemented, it is necessary to assess its performance in the stroke population. The objective of this study was to evaluate the psychometric properties of PROMIS GH in patients with ischemic stroke and intracerebral hemorrhage.
PROMIS GH and 6 PROMIS domain scales measuring same/similar constructs were electronically collected on 1102 patients with ischemic and hemorrhagic strokes at various stages of recovery from their stroke who were seen in a cerebrovascular clinic from October 12, 2015, through June 2, 2017. Confirmatory factor analysis was performed to evaluate the adequacy of 2-factor structure of component scores. Test-retest reliability and convergent validity of PROMIS GH items and component scores were assessed. Discriminant validity and responsiveness were compared between PROMIS GH and PROMIS domain scales measuring the same or related constructs. Analyses were repeated stratified by stroke subtype and modified Rankin Scale score <2 versus ≥2.
There was moderate internal reliability (ordinal α, 0.82-0.88) and marginal model fit for the 2-factor solution for component scores (root mean square error of approximation, 0.11). Convergent validity was good with significant correlations between all PROMIS GH items and PROMIS domain scales (<0.001 for all). There was excellent discrimination for all PROMIS GH items and component scores across modified Rankin Scale levels. Good responsiveness (effect size, >0.5) was demonstrated for 8 of the 10 PROMIS GH items. Reliability and validity remained consistent across stroke subtype and disability level (modified Rankin Scale, <2 versus ≥2).
PROMIS GH exhibits acceptable performance in patients with stroke. Our findings support International Consortium for Health Outcomes Measurement recommendation to use PROMIS GH as part of the standard set of outcome measures in stroke.
国际健康结局测量协会最近将10项患者报告结局测量信息系统全球健康（PROMIS GH）量表纳入其推荐的卒中结局测量标准集。在广泛实施PROMIS GH收集之前，有必要评估其在卒中人群中的表现。本研究的目的是评估PROMIS GH在缺血性卒中和脑出血患者中的心理测量特性。
2015年10月12日至2017年6月2日期间，在一家脑血管诊所对1102例处于卒中恢复不同阶段的缺血性和出血性卒中患者以电子方式收集了PROMIS GH以及6个测量相同/相似结构的PROMIS领域量表。进行验证性因子分析以评估分量表2因子结构的适宜性。评估了PROMIS GH项目和分量表的重测信度和收敛效度。比较了PROMIS GH与测量相同或相关结构的PROMIS领域量表之间的区分效度和反应度。按卒中亚型以及改良Rankin量表评分<2与≥2进行分层重复分析。
分量表具有中等的内部信度（有序α，0.82 - 0.88），且2因子解的模型拟合度一般（近似均方根误差，0.11）。所有PROMIS GH项目与PROMIS领域量表之间均具有显著相关性，收敛效度良好（所有P < 0.001）。在改良Rankin量表各水平上，所有PROMIS GH项目和分量表均具有出色的区分度。10项PROMIS GH项目中有8项显示出良好的反应度（效应量>0.5）。信度和效度在卒中亚型和残疾水平（改良Rankin量表，<2与≥2）之间保持一致。
PROMIS GH在卒中患者中表现出可接受的性能。我们的研究结果支持国际健康结局测量协会关于将PROMIS GH用作卒中结局测量标准集一部分的建议。
