Detecting subgroups in social participation among individuals living with spinal cord injury: a longitudinal analysis of community survey data

STUDY DESIGN: Longitudinal community survey. OBJECTIVES: To determine subgroups in social participation of individuals living with spinal cord injury (SCI). SETTING: Community. METHODS: Data were collected in 2012 and 2017 as part of the community survey of the Swiss Spinal Cord Injury cohort. Participation was assessed using the 33-item Utrecht Scale of Evaluation of Rehabilitation-Participation evaluating frequency of, restrictions in and satisfaction with productive, leisure, and social activities. Linear mixed-effects model trees were used to distinguish subgroups in participation associated with sociodemographic and lesion characteristics. RESULTS: In all, 3079 observations were used for the analysis, of which 1549 originated from Survey 2012, 1530 from Survey 2017, and 761 from both surveys. Participants were mostly male (2012: 71.5%; 2017: 71.2%), aged on average 50 years (2012: 52.3; 2017: 56.5), with an incomplete paraplegia (2012: 37.5%; 2017: 41.8%) of traumatic origin (2012: 84.7%; 2017: 79.3%). There was limited within-person variation in participation over the 5-year period. Participation varied with age, SCI severity, education, financial strain, number of self-reported health conditions (SHCs), and disability pension level. Among modifiable parameters, the number of SHCs and disability pension level emerged as the most frequent partitioning variables, while education was most informative for participation in productive, leisure, and social activities. CONCLUSIONS: Long-term rehabilitation management and clinical practice should target people most prone to decreased participation in major life domains. Our study indicates that the alleviation of SHCs, engagement in further education, or adjusting disability pension level are promising areas to improve participation of persons living with SCI