News headlines such as "Hunger kills 6 million children a year" are commonplace, but many readers are ill-equipped to assess or interpret the quality of the information provided, according to an article in a Series about health statistics in this weeks issue of The Lancet.

Politicians, policymakers, and public-health professionals make complex decisions on the basis of estimates of disease burden from different sources, many of which are "marketed" by skilled advocates. Understanding how closely linked a type of statistic is to disease and death rates is crucial in designing health policies and programmes.

To help people who rely on global public-health statistics make informed decisions, Robert E Black (Johns Hopkins Bloomberg School of Public Health, Baltimore, USA) and colleagues explain how health estimates are developed, and offer basic guidance on how to assess and interpret them. They highlight that estimates would be more credible if they come from technical groups that are independent of the organisations that implement programmes and advocate for funds.

The authors conclude: "What is needed is a combination of normative principles for the development and communication of estimates, combined with efforts to improve the skills of policymakers in interpreting them. Good public-health decisions will always be a shared responsibility between those who generate data and estimates, and those who use them to make decisions A common framework and standardised methods, building on the work of [WHO/UNICEF] Child Health Epidemiology Reference Group and other groups, are urgently needed".