Characterizing Wheelchair-Related Falls in Elderly Veterans

Rationale and Objectives. A multitude of risk factors for falls has been reported, however many of these risk factors are not applicable to persons who use a wheelchair for mobility. Specific objectives include: (1) Describe the incidence and prevalence of wheelchair tips (near falls), falls, and fall-related injuries, (2) Describe the epidemiology of the wheelchair tip or fall event. (3) Create models for predicting wheelchair tips, falls, and fall-related injuries for elderly persons who use a wheelchair for mobility. (4) Determine healthcare utilization and direct costs associated with wheelchair-related fall injuries. (5) Describe patient perceived short- and long-term consequences of wheelchair-related falls.

Research Design. This 3-year prospective cohort study includes qualitative and quantitative data collected at baseline data and through monthly follow-up phone interviews over a 12-month period. Data Source. Baseline data on demographics, intrinsic risk factors, and extrinsic risk factors will be collected through interviews. We will assess functional status, cognition, and home management skills through standardized tools. Administrative databases will be used to gather data regarding comorbidities, Veterans Health Administration (VHA) health care utilization and direct costs. A physical therapy assistant will also conduct a wheelchair inventory and wheelchair skills test. Monthly follow-up phone interviews will include number of falls, description of the fall event/mechanism and nature of the injury, changes in modifiable risk factors, and description of injury and treatment. A separate phone interview will be conducted with a subset of subjects at 6 months post fall, using open-ended questions to gather data about patient-perceived long-term consequences of falls. At the end of the 12-month data collection, we will reassess patient's functional status and cognition. Sample. Inclusion criterion includes all veterans who (1) are aged 62 and older, (2) use a wheelchair for their primary means of mobility, (3) have used a wheelchair for at least 12 months previous to enrollment, and (4) will be using a wheelchair for at least the next 12 months. Exclusion criteria (1) no longer living in a community-based setting (e.g., admitted to nursing home), (2) do not have a telephone. We will over sample women and minorities. A total of 882 subjects are needed for a power of 0.8 and an alpha of 0.05 with an anticipated drop out rate of 30%. Analysis. Bivariate analysis will be performed to determine relationships between outcomes and risk factors. Logistic regression models for predicting wheelchair-related tips, falls and fall-related injuries will be developed based on the most important intrinsic and extrinsic risk factors, controlling for possible confounders. Falls will also be analyzed according to a survival-analysis technique. Univariate and multiple Cox regression will be used to assess the associations of different independent variables with the injurious fall. The cost analysis estimates the net or marginal costs of wheelchair users who experience a wheelchair fall accident and seek medical advice or treatment. We use a dichotomous fall (generating medical expense) variable for a "with" or "without" analysis to compare the average cost of medical care for wheelchair users without a fall involving medical observation or care to the average cost of those that did have a serious fall (including death). The difference between these two groups is a proxy of the cost of utilization due to the fall. Multivariate regression analysis will be used to gauge the net effects of a fall on costs, controlling for the wider array of factors that drive cost of treatment and thereby potentially confound univariate analysis. The control variables are demographic including co-morbid conditions and healthcare related.

Anticipated Impact. Our project is expected to identify previously unaccounted for factors that predispose persons who use a wheelchair to falls and fall-related injuries. We will create models for predicting wheelchair tips, falls, and fall-related injuries. These models will be used to develop an evidence-based, patient centered wheelchair falls prevention program targeting modifiable risk factors. Findings from this study will be so used to develop an instrument to identify fall risk in elderly wheelchair users as well as an evidenced-based prevention program. We will partner with the VA National Center for Patient Safety to disseminate these products nationally in the VHA.