Introduction: The most effective use of trauma center resources helps reduce morbidity and mortality, while saving costs. Identifying critical infrastructure characteristics, patient characteristics and staffing components of a trauma center associated with the proportion of patients needing major trauma care will help planners create better systems for patient care.

Methods: We used the 2009 National Trauma Data Bank-Research Dataset to determine the proportion of critically injured patients requiring the resources of a trauma center within each Level I-IV trauma center (n=443). The outcome variable was defined as the portion of treated patients who were critically injured. We defined the need for critical trauma resources and interventions ("trauma center need") as death prior to hospital discharge, admission to the intensive care unit, or admission to the operating room from the emergency department as a result of acute traumatic injury. Generalized Linear Modeling (GLM) was used to determine how hospital infrastructure, staffing Levels, and patient characteristics contributed to trauma center need.

Results: Nonprofit Level I and II trauma centers were significantly associated with higher levels of trauma center need. Trauma centers that had a higher percentage of transferred patients or a lower percentage of insured patients were associated with a higher proportion of trauma center need. Hospital infrastructure characteristics, such as bed capacity and intensive care unit capacity, were not associated with trauma center need. A GLM for Level III and IV trauma centers showed that the number of trauma surgeons on staff was associated with trauma center need.

Conclusion: Because the proportion of trauma center need is predominantly influenced by hospital type, transfer frequency, and insurance status, it is important for administrators to consider patient population characteristics of the catchment area when planning the construction of new trauma centers or when coordinating care within state or regional trauma systems.