Level-I trauma centre treatment effects on return to work in teaching hospitals

Previous research found a positive effect of Level-I trauma centres on return to work outcomes for patients 18-64 years old who were mainly working before injury. Trauma centres were compared to hospitals that differed on average in characteristics such as size and staffing, among others. Thus, a portion of the effect found could be due to general differences in hospital variables rather than the special characteristics of Level I trauma centres. Comparing Level I trauma centres to other Teaching hospitals provides a more refined test of the effect of these centres on return-to-work outcomes.

The National Study on the Costs and Outcomes of Trauma (NSCOT) is the main source of data for our empirical investigation. We used non-linear instrumental variables methods to control for unobserved characteristics and restrict the sample to teaching hospitals. The first method is the two-stage residual inclusion model in which we identify the effect using the proportion of resident population served by Helicopter Ambulance Services (at the state level) as an instrumental variable. The second method is a recursive bivariate probit model.

We found that treatment at Level-I trauma centres has a positive effect on return to work outcomes three months after injury. The estimated effect is statistically significant and positive, but lower than the estimate that did not focus on teaching hospitals.

A previous study found positive effects of treatment at a Level-I trauma centre on return-to-work outcomes, however, a portion of the effect found was due to general differences in hospital variables.