Unsupported sitting in people with spinal cord injury: the development and validation of tests for assessment and evaluating the effects of therapy

Unsupported sitting is important for people with spinal cord injuries (SCI) to be able to perform everyday tasks independently. The aims of this study were threefold: to develop clinical assessment tools to measure unsupported sitting ability; to develop a scale assessing concern about falling; and to utilize all these measures in a randomized controlled trial (RCT) designed to evaluate the effectiveness of training unsupported sitting in people with SCI. Thirty adults (C6-L1) performed six simple tests on two occasions, including the testing of: timed unsupported sitting; maximal torso leaning; timed donning/doffing of a T-shirt; fast alternating arm reaching; seated upper body/arm reaching; and coordinated stability. All the tests were validated by chronicity and injury level (p<0.05) and demonstrated good to excellent reliability (ICC=0.72-0.91). A scale was developed which addressed concern about falling during 16 activities of daily living associated with falling and specific to people with SCI. In total, 125 adults completed the scale and questions related to their SCI and physical abilities. The scale had excellent internal and test-retest reliability (Cronbach&#146;s &#945;=0.92, ICC=0.93). The scale discriminated between groups likely to have real differences, indicating good validity. Groups showing high levels of concern included people with high level injuries, minimal falling history, dependence in vertical transfers, self-reported fear of falling and poor perceived sitting ability (p<0.05). Reliable and valid measures have now been developed to assess unsupported sitting in people with SCI. These measures are suitable for clinical and research use. A second group of thirty adults with chronic SCI (T1-T12) participated in the RCT. The training group (n=15) performed one hour of task-specific exercise training in unsupported sitting three times per week for six weeks. The control group (n=15) did not receive training. Of the primary outcome measures, the training group showed a marked improvement in maximal torso leaning range (between-group mean difference 64 mm, 95% CI 20-108 mm; p=0.006). Significant improvements were also shown for two secondary outcome measures of reaching (p