Epidemiology and Cost of Falls in Veterans With a Spinal Cord Injury

Overview

Brief Summary Detailed Description Background: A multitude of risk factors for falling has been reported for the elderly, however many of these risk factors are not applicable to persons with a spinal cord injury, who use a wheelchair for mobility. The objectives of this study are to: (1) describe the incidence and prevalence of tips, falls, and fall-related injuries in SCI; (2) describe the epidemiology of the fall event (person, time, place, and activity); (3) describe any injuries associated with a fall, including the mechanism and nature of the injury, severity of injury, as well as treatment required; (4) determine the risk factors related to falls and fall-related injuries in SCI, and develop a model for predicting falls and fall-related injuries in SCI; (5) determine healthcare utilization as well as direct and indirect costs associated with fall-related injuries in SCI; and (6) describe patient-perceived short- and long-term consequences of falls in SCI. The data collection for this study was completed on March 31, 2007, and data analysis is in final phase. Objectives: The objectives of the study are to describe: (1) the incidence and prevalence of wheelchair tips, falls, and fall-related injuries; (2) epidemiology of event; (3) injuries associated with event, including mechanism, nature, severity of injury and treatment required; (4) determine risk factors and develop a model for predicting tips and falls; (5) determine healthcare utilization and direct and indirect costs; and (6) describe perceived short- and long-term consequences of falls in SCI. Methods: This is a prospective cohort study. Data was collected through patient surveys, medical records, and VA databases. Baseline information includes risk factors (e.g., user characteristics, wheelchair features, wheelchair activities, etc.) and physical environmental data. Monthly follow-up calls tracked tips, falls, and injuries. Status: Recruitment ended on April 1, 2006 with a total of 702 subjects, and data collection was completed on March 31, 2007. The Data Safety Monitoring Board for this study conducted its final meeting on September 17, 2007. Data has been analyzed and 17 manuscripts are currently in development. Impact: While much is known about falls in elderly, there is a lack of understanding of the epidemiology of wheelchair-related falls. Our project is expected to identify previously unaccounted for factors that predispose persons with SCI to falls and fall-related injuries. This study will result in the creation of a model for predicting falls and fall-related injuries in SCI. The predictive model will be used to develop intervention strategies targeting modifiable risk factors. Eventually, we will test this predictive model with other vulnerable veteran populations. Findings from this study wil be used to develop an instrument to identify fall risk in persons with SCI.

Of the 702 patients enrolled, 659 subjects provided data for follow up analysis. The primarily male (96%) population was 55 + 13 years old, and had their SCI for an average of 21 + 13 years. 47% of the injury levels were cervical and 46% thoracic. The majority of subjects used manual wheelchairs (62%) an average of 10.9 + 4.3 hours a day. 553 fall events were reported by 204 individuals. 95 (14%) subjects were injured as a result of the fall. Regression models showed that fall occurrence was predicted by increased pain in past two months, increased alcohol use, higher FIM sub-motor scores, having a previous fall, shorter wheelchair length, fewer years of SCI, and not having one's first wheelchair. Injurious falls were predicted by increased pain in past two months, higher FIM sub-motor scores, having a previous fall, non-flat home entrance and a diagnosis of arthritis. In both models, 83% of the variance was explained by the predictors. Additional analysis is planned for completion of manuscripts. In addition, funding was received to develop a risk assessment tool for falls from wheelchairs in SCI. Final analyses are being conducted and 17 manuscripts are currently in development.