The Global Burden of Disease and Injury project (GBD) is a monumental undertaking. First commissioned by the World Bank in 1990, and more recently supported by the Bill and Melinda Gates Foundation, this study aims to quantify the global burden and distribution of mortality, morbidity and disability attributable to a comprehensive set of diseases, conditions and risk factors. Results are most commonly presented as disability adjusted life years (DALYs), a common metric that accounts for years of life lost and for years of life lived with pain, impairment or other disability. Causes are grouped into three large domains: (1) communicable, neonatal, maternal and nutritional disorders; (2) non-communicable diseases and (3) injuries.

The methods employed to generate these estimates are both state-of-the-art and daunting, involving huge human and computational resources.1 The raw data are amassed from vital registries, community and household surveys, hospital discharge abstracts and a host of other sources, including published epidemiological studies. These data are fed into sophisticated modelling programmes that generate global, regional, national and—in some cases—subnational—estimates, along with appropriate uncertainty intervals. In some cases this means extrapolating assumptions from one country or region to another. In others, it means glossing over the important—but unmeasured—differences that occur within countries and among demographic subpopulations.

These methods have been presented by the authors in multiple venues over the years and, while not without controversy, are generally seen as the best available …