In a mature trauma system, there is no difference in outcome (survival) between Level I and Level II trauma centers.

Background: The state of Pennsylvania (PA) has one of the oldest, most well-established trauma systems in the country. The requirements for verification for Level I versus Level II trauma centers within PA differ minimally (only in the requirement for patient volume, residency, and research). We hypothesized that there would be no difference in outcome at Level I versus Level II trauma centers.

Methods: Odds of mortality for 16 Level I and 11 Level II hospitals in PA over a 5-year period (2004-2008) was computed using a random effects logistic regression model. Overall adjusted mortality rates at Level I versus Level II hospitals were compared using the nonparametric Wilcoxon's rank sum test. The crude mortality rates for 140,691 patients over the 5-year period were similar (5.07% Level II vs. 5.48% Level I), but statistically significant (odds ratio mortality at Level I = 1.084, p = 0.002 Fisher's exact test).

Results: Although Level I centers had on average crude mortality rates that were higher than those of Level II centers, median adjusted mortality rates were not different for the two types of centers (Wilcoxon's rank sum test). Performance of Level I versus Level II shows considerable variability among centers (basic random effects model, age, blunt/penetrating, and Injury Severity Score [ISS]). However, Level II centers seem no different from Level I.

Conclusion: As trauma systems mature, the distinction between Level I and Level II trauma centers blurs. The hierarchal descriptors "Level I" or "Level II" in a mature trauma system is pejorative and implies in those hospitals labeled "Level II" as inferior, and as such should be replaced with nonhierarchal descriptors.