Purpose The objective of this study is to estimate the effect of community-level and individual-level risks on hospitalization of older populations after traffic accidents.

Method This is a case-control study using two different sets of data were used in this study: Korean National Hospital Discharge In-depth Injury Surveillance data from 2012 to 2014 as case and Community Health Survey (CHS) data from 2012 to 2014 as control. Variables for individual risks were gender and age. For community risk factors, community rate of traffic accident occurrence at night, at cross-walks and at intersections were used. Multilevel regression models were used to analyze the relationship between community risk factors and injury hospitalization.

Result Total 3847 patients 65 years or older hospitalized for traffic accidents were defined as cases and 15,932 CHS respondents were defined as controls. Compared with non-hospitalized populations, the odds ratio of male for hospitalization in older population was 0.482 (95% CI 0.445 to 0.521) compared to female and odds ratio for unit change of age was 0.974 (95% CI 0.967 to 0.981). After controlling individual-level variables, community night traffic accident rate showed odds ratio of 1.055 (95% CI 1.021 to 1.09), community rate of cross-walk accident 0.927 (95% CI 0.850 to 1.012) and rate of intersection accident showed 1.021 (95% CI 1.003 to 1.038).

Conclusion Both individual-level and community traffic accidents status influenced the risk for hospitalization of older population. To reduce the burden of mortality and morbidity from traffic accidents in older people, we should look at not also individual risks but also community-based problems. Further studies are needed to establish concrete strategies.