The Spinal Cord Injury Pressure Ulcer Scale (SCIPUS): an assessment of validity using Rasch analysis

• Published in: Spinal cord 2019-10, Vol.57 (10), p.874-880
• Main Authors: Higgins, Johanne, Laramée, Marie-Thérèse, Harrison, Kate Rousseau, Delparte, Jude J., Scovil, Carol Y., Flett, Heather M., Burns, Anthony S.
• Format: Article
• Language: English
• Subjects: 692/499; 692/700; Adult; Aged; Anatomy; Biomedical and Life Sciences; Biomedicine; Cross-Sectional Studies; Female; Human Physiology; Humans; Injuries; Injury analysis; Male; Males; Middle Aged; Neurochemistry; Neuropsychology; Neurosciences; Pressure; Pressure Ulcer - etiology; Pressure ulcers; Psychometrics - instrumentation; Rasch model; Rehabilitation; Retrospective Studies; Risk Assessment - methods; Secondary analysis; Spinal cord injuries; Spinal Cord Injuries - complications
• ISSN: 1362-4393; 1476-5624
• DOI: 10.1038/s41393-019-0287-z

Study design Secondary analysis of retrospective data. Objective The aim of this study was to further validate the Spinal Cord Injury Pressure Ulcer Scale (SCIPUS) using Rasch analysis. Setting Two rehabilitation centers in Canada. Method Data were collected as part of the Spinal Cord Injury Knowledge Mobilization Network (SCI KMN) initiative. The SCIPUS was completed within 72 h of inpatient admission. Persons admitted for initial rehabilitation in two inpatient spinal cord rehabilitation programs were included in the project. Results Data from 886 participants were analyzed, approximately 60% of whom were males. Rasch analyses demonstrated that the SCIPUS, in its current format did not meet criteria required for true measurement. A transformed version of the SCIPUS obtained by deletion of misfitting items and modification of the response scales improved fit to the model and showed preliminary evidence of unidimensionality. The person separation index, however indicated that the scale requires further adjustments of its scoring options. Conclusions In its original form, the SCIPUS does not meet the requirements of the Rasch model and its total score should be used cautiously. However, following some adjustments to the items such as addressing DIF between sites to insure a standardized assessment across sites and adding response options to some of the items, interval-scale measurement should be possible.