Traumatic injury represents a leading cause of mortality and loss of independence among the growing portion of the US population aged >64y. In 2013, nearly 4 million non-fatal traumatic injuries occurred among older patients. By 2050, the number of individuals aged >64y is projected to double with an anticipated parallel increase in the number of older trauma patients. Traumatic brain injury (TBI) (>425,000/anum) and hip fracture (>258,000/anum) are two of the most common and debilitating injuries occurring among older adults, often resulting in long-term functional impairments, nursing home admission, decreased independence, and shortened life expectancy. Ensuring that older patients receive the optimal level of trauma center (TC) care may be one important way to improve outcomes. While TC care has been shown to significantly reduce mortality in young, severely-injured populations, its effectiveness in improving outcomes among older patients remains unclear. This proposal aims to compare the effectiveness of TC versus non-trauma center care in older patients. We will use 2007-2014 Medicare claims data, along with data from the American Trauma Society, American Hospital Association, and University of Pennsylvania Cartography Modeling Lab (geographic assessment), to understand the unique care patterns and management needs of older injured patients. We will first determine where older patients receive trauma care and identify characteristics (demographic, injury, geography-related) associated with treatment at different TC levels using descriptive statistics and ordered logistic regression. Calculation of injury and comorbidity measures will be conducted through the adaptation and validation of a SAS- compatible version of the ICDPIC program written for STATA to enable use with the most up-to- date Medicare information available through CMS's Virtual Research Data Center. We will then examine possible effects of TC care on outcomes among older patients with traumatic injury, comparing differences in mortality, morbidity, LOS, discharge disposition, readmission, cost, and functional status between patients treated at the highest versus lowest levels of trauma care, at discharge and at 30, 90, and 365 days. Finally, we will identify possible heterogeneity in the effectiveness of TC care and estimate the potential impact of treating older patients at the facility most appropriate for their injury, using multiple interaction proportional hazard modeling, accelerated failure time modeling, and counterfactual techniques (Markov chain/random field modeling). Variations in costs will be explored to determine possible cost saving. Throughout all of these analyses, we will identify subsets of older injured patients who benefit from TC care, specifically focusing on differences between older adults with TBI (expected to benefit from high-level TC care) and hip fracture (thought to be manageable regardless of treating facility).
Trauma is a major cause of morbidity and mortality among Americans aged >64y. Ensuring that older patients receive the optimal level of trauma center care may be one important way to improve outcomes; however, the effectiveness of trauma centers in improving outcomes among older patients remains unclear. This proposal will determine the effectiveness of trauma center care among older patients, specifically focusing on differences between traumatic brain injury (expected to benefit from higher-level trauma center care) and hip fracture (manageable regardless of treating facility) while addressing short- and long-term outcomes including mortality, complications, readmissions, functional status, and cumulative direct costs.
