Association between socioeconomic status and burn injury severity.

BACKGROUND: It has been previously established that the incidence of burn is higher in lower socioeconomic (SES) groups. What is not clear, however, is whether or not the severity of burn is also higher in lower SES groups. The purpose of this study is to establish a relationship between household-level socioeconomic status (SES) and severity-based incidence of burn.
METHODS: A burn injury database was generated from the National Injury Database (2001-2003) with a 1-year follow-up period containing information about the date and time of burn injury, the International Classification of Disease 10th Edition-based (ICD-10) diagnostic codes, gender, age, residence, and type of insurance. In addition, we calculated the severity of each burn using the Excess Mortality Ratio-adjusted Injury Severity Score (EMR-ISS), which is similar to the New Injury Severity Score. Socioeconomic status was measured on the basis of quintiles of premiums for National Health Insurance, which was decided on a household-level either by monthly salary (Employee Insurance, EI) or by owned property (Self-Employed Insurance, SEI). Medical Aid (MA) population was regarded as a reference. After calculation of 5 year-gender specific standardized incidence rates (SIRs) of burn by SES groups, the association of SES and severity of burn was evaluated using a multivariate logistic regression model and the Cox-proportional hazard regression analysis.
RESULTS: A total of 870,411 burn cases were examined. The standardized incident rates (SIRs) of mild(1<or=EMR-ISS<or=8), moderate(9<or=EMR-ISS<or=24), severe(25<or=EMR-ISS<or=74), critical(EMR-ISS=75 or death), and total burn injury were 4951, 951, 124, 20, and 6046 per 1,000,000 person-years, respectively. Meaningful changes of relative risk by severity were found: it was greater in higher SES groups (0.67 in the highest SEI, 0.58 in the highest EI) compared to the Medical Aid (MA) population. For injuries classified to be greater than severe (25<or=EMR-ISS), odds ratios according to increase of SES level were significantly decreased to 0.77 in SEI and 0.63 in EI. For critical injuries, hazard ratios were also significantly decreased to 0.51 in the highest SEI and 0.32 in the highest EI.
CONCLUSIONS: Severity-based incidence of burn was significantly affected by household-level SES in a nationwide cohort study, with more severe injuries noted in the lower socioeconomic groups.