THE YEAR 2006 WILL MARK THE 40TH ANNIVERSARY OF the establishment of the first civilian trauma centers in the United States, at San Francisco General Hospital and at Cook County Hospital in Chicago. The same year also marks the 30th anniversary of the American College of Surgeons’ publication of the “Optimal Hospital Resources for Care of the Seriously Injured” document. The parent document and subsequent revisions articulate expert opinion on what constitutes categorization of trauma centers, the resources required, and how trauma centers fit into a system of care for the injured patient. The article by MacKenzie and colleagues in this issue of THE JOURNAL provides a national inventory of hospital trauma centers. The study encompasses the last decade and updates the report published in 1995. The findings are noteworthy and have potentially important implications for health care policy. MacKenzie et al correctly point out that trauma centers are effective, but they do not cite one of the more interesting articles that substantiates studies in the medical literature. According to a report by Harris et al, since 1931 the US homicide rate has not strayed more than 51% from its level of 9.2 per 100000 population. In 1998, the homicide rate was 6.8 per 100000 population, 26% lower than the 1931 level. Data from the Federal Bureau of Investigation National Uniform Crime Report corroborate these levels but also show that aggravated assault rates have increased by 700% since 1931, as have rates of other violent crimes similar to those for aggravated assault. Harris et al attributed this paradox between steady homicide rates and increasing aggravated assault rates to the development of trauma systems. They state, “The homicide paradox involves the observation of parallel, dramatic developments in medical technology and related support services, developments that may have functionally, and equally dramatically, suppressed the homicide rate compared to what it would be had such progress not been made.” They further state, “Compared to 1960, the year our analysis begins, we estimate that without these developments in medical technology, there would have been between 45,000 and 70,000 homicides annually the past five years instead of the actual 15,000 – 20,000.” It is sad that violent crime has increased, but gratifying that trauma centers may have kept homicide rates low. In the current inventory of trauma centers, MacKenzie et al also raise the issue of the optimal number of level I and level II trauma centers. The authors point out that this number should be determined by evidence-based guidelines, and they specifically mention that concentrating severely injured patients into a limited number of specialty trauma care facilities will lead to greater experience with trauma care at these centers and thus improve patient outcomes. A secondary argument for limiting trauma centers is accountability for costs. The 1979 version of the optimal criteria recommended that trauma centers should be limited and based on need. The rationale for this recommendation is that designation of trauma centers by state emergency medical services (EMS) agencies is essentially a franchise. In addition to improving care and outcomes, trauma centers must be accountable for costs to the state legislatures that have granted them the franchise. Trauma centers are perhaps the only example of tertiary care in which the physicians providing the care have made an earnest attempt to limit the number of hospitals based on need, but this has not been uniformly successful. Previous work by Bazzoli and MacKenzie that cataloged the key characteristics of state trauma systems found that determination of need was a main characteristic lacking in a number of these states. Although this current inventory does not look at these key characteristics, the problem still remains in some of the 35 states that have a trauma system. The authors also point out that since 1991, the total number of trauma centers has more than doubled, increasing from 471 to 1154, and the number of level I and level II centers has increased from 374 to 453. However, several problems remain, including geographic distribution in underserved areas, particularly in rural communities. The authors have not addressed how often an inventory should be conducted, or several other components of a trauma system. Such an inventory should be done annually, and the infor-