A Rehabilitation Data Collection Study in Patients With Spinal Cord Injury

This is a prospective, single-center observational registry study conducted at the National Traffic Injury Rehabilitation Hospital in Korea. It forms the first stage of a multi-year initiative (2024-2026, funded by the National Rehabilitation Center) intended to design and establish a comprehensive rehabilitation registry for patients with spinal cord injury (SCI) and to generate foundational data for predicting patients' long-term functional status.

Eligible patients are enrolled during inpatient rehabilitation using a systematic selection approach, after the purpose and procedures of the study-including the follow-up surveys-are explained and written informed consent is obtained. Recruitment continues from study initiation in 2024 until the target enrollment is reached or the study ends in 2026. Data are collected in three waves: a baseline survey at the time of admission, administered by an internal researcher by telephone or in person; and two follow-up surveys, conducted at approximately 6 and 12 months after discharge through interviews performed by a professional research organization. Across these waves, the registry captures demographic and injury-related characteristics, use of health services, health and functional status, and clinical information; the follow-up waves additionally capture post-discharge health, healthcare utilization, employment, household income, and social participation. Data are drawn from a combination of patient-reported responses, physician-collected information, and abstraction from medical records. (Eligibility criteria and specific outcome measures are provided in the corresponding sections and are not repeated here.)

Registry procedures and quality factors:

• Documentation and data dictionary: The full data-collection process is documented in advance, including the protocol, operations and management policies, procedures, data-collection instruments, a complete list of data elements with their definitions, the screening process, the method for generating unique patient registration numbers to prevent duplicate enrollment, and training for data collectors.
• Roles and responsibilities: Personnel involved in data collection and their specific responsibilities are defined, with clear role assignment.
• Data validation and quality assurance: Completed questionnaires are reviewed immediately for missing responses or errors, with re-confirmation from the participant when needed, and data are captured using double data entry to ensure accuracy.
• Data checks: A data management manual is used to review missing values, incorrectly entered values, out-of-range values, and logically impossible values.
• Source data: Each data element is characterized by whether it derives from a primary or secondary source and whether it is patient-reported, physician-collected, or extracted from medical records.
• Data storage and security: Identifiable information (e.g., name, resident registration number, patient number) is anonymized and replaced with randomly generated codes, and only coded files are used for analysis. Original data are stored with two-level encryption (file- and sheet-level) on encrypted media kept on a password-protected computer in a locked room, with documented deletion/modification history, a backup plan, and a security plan.
• Sample size: The planned sample size is 250 participants. A power analysis (G*Power; medium effect size 0.60, two-sided α = 5%, power ≥ 95%) established a required minimum of approximately 100 evaluable participants; enrollment is set at 250 to account for an estimated 20% attrition and to preserve data quality by offsetting loss from missing data.
• Missing data: Missing, out-of-range, and inconsistent values are addressed through the data management manual, and the planned enrollment of 250 participants helps limit reductions in the analyzable sample due to attrition and missing data.
• Statistical analysis: Collected data are summarized using descriptive statistics. Comparative analyses (t-tests, analysis of variance, and chi-square tests) examine differences by characteristics such as sex, age group, and injury type-for example, in rehabilitation service utilization and timing of home return at baseline, and in post-discharge healthcare utilization and health outcomes at follow-up.