Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)

Graduate Group

Epidemiology & Biostatistics

First Advisor

Douglas J. Wiebe


Approximately 3 million Emergency Department (ED) visits and 50,000 deaths occur annually in adults over age 65, of which 50% are from falls and 13% are from motor vehicle crashes. Treatment begins in the out-of-hospital setting (Emergency Medical Services), continues in the ED, includes definitive in hospital and outpatient care, and then recovery in rehabilitation centers and via home-health providers. Older adults have four times the odds of dying in the hospital and are twice as likely to be discharged to skilled nursing facilities when compared with younger patients. The focus of this study is to evaluate the intensity of care delivered to older adults during hospitalization. We used Centers for Medicare and Medicaid Services claims (2013-2014), to identify beneficiaries with moderate and severe blunt trauma admitted through an ED (n=683,398). First, we classified beneficiaries into low, moderate, or high intensity care using latent class methods which examined 18 procedures/interventions, ICU length of stay, and demographic and injury characteristics. Approximately 73% were classified as low intensity, 23% moderate, and 4% as high intensity care. Lower intensity aligned with increased age and reduced injury severity, while males, non-whites, and non-fall mechanisms were more common with high intensity. Second, at each hospital we calculated the average probability that patients received high intensity care (0.20; IQR:0.15-0.26). High intensity hospitals were defined as those with an average probability >0.50 (n=77), and a greater proportion of these were non-trauma centers (93%) and located in the South (52%). Third, we examined the association between survival and intensity and found decreased odds of 30-day survival in moderate and high intensity when compared with low (OR:0.35 (95% CI:0.34, 0.36) and OR:0.07 (0.07, 0.07), respectively). Using a subdistribution hazards model to estimate survival to discharge (competing risk: death), moderate and high intensity both had decreased survival compared to low intensity (SHR:0.56 (95% CI:0.56, 0.57) and 0.21 (0.20. 0.21), respectively). This work demonstrates that although heterogeneous, care for blunt trauma can be evaluated using a single novel measure. Care intensity is related to resources and patient preferences, is a component of quality, and should be benchmarked along with clinical outcomes.

Included in

Epidemiology Commons