Can SCI Databases Inform Clinical Care? Identifying Gaps Between Clinically Relevant Variables and Those Captured in Rehabilitation Data Systems

Abstract Study Design Exploratory, hypothesis-generating mixed‑methods study combining qualitative thematic analysis and quantitative survey ratings. Objectives Identify variables clinicians consider most important for individualized rehabilitation for adults with spinal cord injury (SCI) and compare these priorities with variables in major SCI rehabilitation datasets, highlighting gaps that may limit predictive modeling. Setting SCI rehabilitation clinicians working in inpatient, outpatient, and community settings across Canada, with comparisons to major North American SCI rehabilitation datasets. Methods Semi-structured interviews with three clinicians were analyzed using inductive thematic analysis. Clinicians rated the importance of 58 variables for SCI rehabilitation planning. Variables were compared with the Spinal Cord Injury Model Systems (SCIMS) database and the Canadian Rick Hansen Spinal Cord Injury Registry (RHSCIR), assessing clinical importance, dataset availability, and feasibility of capture. Results Twenty-two clinicians were surveyed (n=19) and interviewed (n=3). Eighteen high-priority variables and five qualitative themes emerged: clinical factors; patient motivation and engagement; caregiver support and environmental readiness; funding/system barriers; and goal alignment. Patient motivation and goals received the highest priority rating, followed by neurological injury level, paraplegia/tetraplegia status, and the American Spinal Injury Association (ASIA) Impairment Scale (AIS) grade. Family supports, home accessibility, living conditions, and discharge destination were also rated highly. Only 6 of the 18 high-priority variables were adequately captured by SCIMS or RHSCIR. Conclusions Clinicians rely on multidimensional variables beyond existing SCI rehabilitation datasets. Psychosocial, environmental, and system-level factors remain underrepresented despite their perceived importance. Incorporating these factors into data collection and predictive models may improve personalized SCI rehabilitation planning.