The recent article by Rosen and colleagues1 presents a sobering reminder that insurance does matter for health, precisely in situations in which the health of patients is most in jeopardy, such as trauma. Unfortunately, as is so often the case when different disciplines fail to interact, the authors missed an opportunity to strengthen support for their findings. They overlooked a carefully designed 2005 study of automobile crash victims in which at least 1 other fatality occurred.2 In the 2005 study, Doyle used detailed data on the nature of automobile crashes linked to hospital discharge data and mortality status for automobile crashes in Wisconsin. Doyle demonstrated that mortality rates were 1.5 percentage points higher for uninsured accident victims compared with insured accident victims, a 40% increase over the mean mortality rate of 3.8%. The uninsured victims also received 20% less medical care (measured by length of stay and spending) than insured comparisons. The mortality estimate in the Doyle study is nearly identical to that in the Rosen et al study (Doyle's estimate would yield an odds ratio of 1.4, the same found in the Rosen et al study). The clinical richness of the current study nicely complements the careful effort to remove confounding factors in the prior study. Together, these and other evidence referenced in the article argue more strongly than ever for the value of health insurance coverage, not simply for financial protection, but to improve health.
Meara E. More Bad News for the Uninsured. Arch Surg. 2010;145(6):604. doi:10.1001/archsurg.2010.100