Trends in the epidemiology of major burn injury among hospitalized patients: a population-based analysis

Background — Burn-related mortality has decreased significantly over the past several decades. While often attributed in part to regionalization of burn care, this has not been evaluated at the population level.

Methods — We conducted a retrospective, population-based cohort study of all patients with > 20% total burned surface area (TBSA) burn injury in Ontario, Canada. Adult (>16y) patients injured between 2003-2013 were included. Deaths in the emergency department were excluded. Logistic generalized estimating equations were used to estimate risk-adjusted 30-day mortality. Mortality trends were compared at burn and non-burn centers.

Results — 772 patients were identified at 84 centers (2 burn, 82 non-burn). Patients were 74% (n=570) male, of median age 46 (IQR 35-60) years and median TBSA 35% (IQR 25-45). Mortality at 30 days was 19% (n=149). The proportion of patients treated at a burn center increased from 57% to 71% between 2003-2013 (p=0.07). Average risk-adjusted 30-day mortality rates decreased over time; there were significantly reduced odds of death in 2010-2013 compared to 2003-2006 (OR 0.39, 95% CI 0.25-0.61). Burn centers exhibited significantly reduced mortality from 2003-2006 to 2010-2013 (OR 0.36, 95% CI 0.34-0.38) compared to non-designated centers (OR 0.41, 95% CI 0.13-1.24).

Conclusions — Mortality rates have decreased over time; significant improvements have occurred at burn centers while mortality rates at non-designated centers vary widely. A high proportion of patients continue to receive care outside of burn centers. These data suggest there are further opportunities to regionalize burn care and in so doing, potentially lower burn-related mortality.

Level of Evidence — Level III epidemiological study