脑卒中PRO量表纸质化和电子化测量方式的等同性分析

Analysis of Equivalence between Paper-based and Electronic Patient-reported Outcome Measures for Stroke

目的对于自行研发的电子化脑卒中患者报告结局(electronic patient-reported outcome,ePRO)量表,考核量表的等同性,即条目功能差异(differential item functioning,DIF),旨在比较纸质化和电子化的测量方式是否影响受试者对PRO量表的应答。方法使用自行研发的App采集患者信息,使用均值协方差结构(mean and covariance structure,MACS)模型考查DIF,分析PRO量表管理方式由纸质化转变为手机App后,量表是否保持了良好的测量等同性。结果量表的DIF考核结果如下:条目PHD(physical domain,生理领域)7、PHD8、PHD14、PSD(psychological domain,心理领域)3、PSD9、SOD(social domain,社会领域)1、SOD6均存在一致性DIF(P 1<α1),条目PHD9、PHD10、PHD11、PHD12、SOD5、THD(therapeutic domain,治疗领域)1存在非一致性DIF(P 2<α2),条目SOD7和THD2同时存在两种形式DIF。以上15个条目存在测量方式间的DIF,但尚在可接受范围。结论本文研发的移动的患者报告结局(mobile-based patient-reported outcomes,mPRO)随访系统,通过对量表不同数据收集方式间的DIF进行考核,证实了电子化的数据收集方式并不影响受试者的应答,测量方式的转变未影响量表测量等同性。该系统得到的测量结果准确可靠,可在临床实践中使用,并在应用过程中对其进行优化。

Objective For the development of the electronic patient-reported outcome measures(ePRO)for stroke,the scale′s equivalence,i.e,differential item functioning(DIF)was assessed,aiming to compare whether paper-and mobile-based collection methods affect the participants′responses.Methods The patient information was collected using the developed App,and the DIF was detected using the mean and covariance structure(MACS)model to analyse whether the scale maintained a good measurement equivalence after changing to mobile-based from paper-based.Results Items PHD7,PHD8,PHD14,PSD3,PSD9,SOD1,and SOD6 exhibited uniform DIF(P_(1)<α_(1)),PHD9,PHD10,PHD11,PHD12,SOD5,and THD1 exhibited non-uniform DIF(P_(2)<α_(2)),and SOD7 and THD2 exhibited both uniform and non-uniform DIF.The DIF between the different measurement modalities existed for the above 15 items;however,it was within the acceptable limit.Conclusion The mPRO patient follow-up system developed in this study by assessing the DIF between the scale′s different data collection methods confirmed that mobile-based data collection did not affect the participants′response,and the measurement method shift did not affect the scale measurement equivalence.The system yielded accurate and reliable measurements that can be used in clinical practice and optimized in its application course.