Psychometric properties of the Swedish versions of Spinal Cord Independence Measure IV (SCIM IV) and Self-report (SCIM-SR) in inpatient and outpatient rehabilitation settings.

Abstract

Psychometric study. To evaluate the data completeness, data distribution and ceiling/floor effects, internal consistency and convergent validity of the Swedish versions of the Spinal Cord Independence Measure IV (s-SCIM IV) and the Spinal Cord Independence Measure Self-report (s-SCIM-SR). Swedish inpatient and outpatient spinal cord injury (SCI) rehabilitation. The translation process was based on established guidelines with researchers, clinicians and consumers. s-SCIM IV and FIM^TM assessments were performed by observation and/or interview, s-SCIM-SR through self-report using paper forms. In total, 101 participants (82% men) were included. There were no missing data for s-SCIM IV and 92% had answered all items in s-SCIM-SR. No ceiling or floor effects were observed. Cronbach´s alpha for the total s-SCIM IV scale was 0.91 (subscales 0.68-0.93) and for the total s-SCIM-SR scale 0.91 (subscales 0.62-0.93), with the lowest alphas for Respiration and Sphincter Management in both outcome measures. s-SCIM IV and s-SCIM-SR correlated strongly with each other and with FIM^TM. Our results support the data completeness, lack of ceiling/floor effects, internal consistency (except the Respiration and Sphincter Management subscale) and convergent validity of the s-SCIM IV and s-SCIM-SR. Based on this initial psychometric testing, these outcome measures can be considered suitable to assess physical independence in inpatient and outpatient rehabilitation and long-term follow-up after SCI, for both clinical and research purposes. The now available Swedish versions of SCIM will enable a uniform national assessment of SCI-specific physical independence and facilitate research and international collaborations and comparisons.