Identification of classes of functioning trajectories and their predictors in persons with spinal cord injury attending initial rehabilitation in Switzerland

Abstract Objective To identify classes of functioning trajectories in persons with spinal cord injury (SCI) undergoing initial rehabilitation following injury, and to examine potential predictors of class membership in order to inform clinical planning of the rehabilitation process. Design Longitudinal analysis of the subject' rehabilitation stay using data from the Inception Cohort of the Swiss Spinal Cord Injury Cohort Study (SwiSCI). Setting Initial rehabilitation in specialized centres in Switzerland. Participants Individuals with newly acquired SCI who completed initial rehabilitation between May 2013 and September 2019. Within this study, 748 participants were included which had a mean age of 54.66 years (SD=18.38 years). The cohort was primarily comprised of males (67.51%) with incomplete (67.51%) paraplegia (56.15%) and a traumatic injury (55.48%). Interventions Not applicable. Main Outcome Measures Functioning was operationalized with the interval-based sum score of the Spinal Cord Independence Measure version III (SCIM III). For each individual, the SCIM III sum score was assessed at up to four time points during rehabilitation stay. The corresponding time of assessment was recorded by the difference in days between the SCIM III assessment and the admission to rehabilitation programme. Results Latent process mixed model (LPMM) analysis revealed four classes of functioning trajectories within the present sample (N=748): Class-specific predicted mean functioning trajectories describe stable high functioning (N=307, 41.04%), early functioning improvement (N=39, 5.21%), moderate functioning improvement (N=287, 38.37%), and slow functioning improvement (N=115, 15.37%), respectively. Out of twelve tested factors, multinomial logistic regression showed that age, injury level, injury severity, and ventilator assistance are robust predictors that can distinguish between identified classes of functioning trajectories in the present sample. Conclusions The study establishes a foundation for future research on the course of functioning of persons with SCI in initial rehabilitation by identifying classes of functioning trajectories. This supports the development of specifically tailored rehabilitation programmes and prediction models, which can be integrated into clinical rehabilitation planning.