Rasch Analysis of the Qualifier Response Options of the ICF Set for Spinal Cord Injury Nursing

李琨,燕铁斌,尤黎明,谢粟梅,李云,唐洁,王颖敏,高焱
DOI: https://doi.org/10.3760/cma.j.issn.0254-1424.2017.02.007
2017-01-01
Abstract:Objective To test the suitability of the qualifier response options of the Body Functions (BF) component of the ICF Set for Spinal Cord Injury Nursing (ICF-SCIN) using Rasch analysis.Methods A total of 140 spinal cord injury patients were recruited between December 2013 and March 2014 from four research centers.Nurses used the BF component of the ICF-SCIN to rate their functioning.Rasch analysis was performed using RUMM 2030 software.Rasch model fit,threshold estimates,and person separation index (PSI) were quantified.Results Twenty-six categories (78.8%) displayed disordered thresholds among the level 1,2 and 3 qualifiers.After rescoring the qualifiers from 0,1,2,3 and 4 to 0,1,1,1 and 2,the threshold estimates of the ICF categories showed a gradually increasing tendency and the BF component demonstrated good fit to the Rasch model (x2=93.96,P ≤ 0.05).The mean fit residuals (standard deviations) for items and persons were-0.14 (0.94) and-0.35 (0.98).The PSI was 0.82.Conclusions The categories in the BF component of the ICF-SCIN showed increased Rasch model fit and an increasing tendency of threshold ordering.The overall model of the component was also improved.The standards for quantifying the ICF qualifiers should be more explicit in future studies.
What problem does this paper attempt to address?