ABSTRACT
BACKGROUND: Age, total burn surface area (TBSA), and inhalation injury are proven predictors of mortality and morbidity following burn injury. Most previous studies have also found that African Americans and females with burns also fare worse. We sought to determine whether these disparities were reduced when burn victims were analyzed separately by categories of insurance coverage. METHODS: We evaluated records in the National Burn Registry (NBR) from 2002 to 2011. Multivariate logistic regression was performed to determine factors associated with inpatient mortality, including age, TBSA, inhalation injury, race, and sex, and allowing for clustering by hospital. Separate models were constructed for each category of insurance. 95% confidence intervals (CI) not including 1 for any odds ratio were considered evidence of statistical significance (designated by * in the table below). RESULTS: NBR included records from 172,640 patients (55.8% Caucasian, 18.1% African American, 14.2% Hispanic, 6.4% other minority groups, 5.4% unknown). Age, TBSA, and inhalation were strong predictors of mortality as expected. Non-African American males were the largest group for all insurance categories, and had the lowest mortality. Controlling for these factors, and compared with non-African American males, African American males had consistently increased odds of mortality regardless of insurance coverage. African American females had increased odds of mortality if they had Private, Medicare, or Medicaid insurance, and Non-African American females had increased odds of mortality if they had Private or Medicaid insurance. The association of Hispanic ethnicity with mortality was inconsistent or insignificant, and other minority groups had too few members to evaluate. Most patients were missing comorbidity data, and no other socioeconomic or hospital data were available in NBR. CONCLUSIONS: African American males with burn injury are at increased risk of mortality regardless of insurance coverage, and most females are at increased risk regardless of race. Analyzing the reasons for these disparities will require databases containing more complete comorbidity, socioeconomic, and/or hospital data.
