Comparing road safety performance across countries: Do data source and type of mortality indicator matter?

This study examined the impact of data source estimates (World Health Organization (WHO) versus Global Burden of Disease (GBD)) and the type of mortality indicator (population-based versus exposure-based mortality) on road safety performance evaluation. Data were derived from WHO publications and the GBD results tool, and we calculated mortality rate ratio (MRR) and differences in country ranking between the two data sources, plus differences in country rankings and in mortality changes between 2010 and 2013 for population-based and vehicle-based mortality. Of 172 countries in both datasets, 32 countries (19%) had low consistency across the two data sources (MRR ≤ 0.49 or ≥1.51). Using population-based mortality data to rank the 172 countries, 77 (45%) had ≥ 20 position difference between the two data sources. Population-based vs. vehicle-based mortality data yielded ≥ 20 position difference in 33 countries for WHO estimates and 42 for GBD estimates. Among the 80 countries having comparable population-based and vehicle-based GBD mortality rates over time, 9 countries displayed opposite changing directions - that is, the change increased in one mortality indicator but decreased in the other indicator between 2010 and 2013. Data source and type of mortality indicators yield a substantial impact on ranking road safety performance across countries, as they are widely used for decision-making by global and national policy-makers and injury researchers. The differences between WHO and GBD estimates may arise from inconsistencies in data input and estimation models. Exposure-based indicators should be preferred in road safety evaluation when data are available. Advanced research is needed to interpret large country variations in road traffic mortality and mortality progress and to develop strategies to narrow the gaps across countries.