If your birth is not registered then you don’t exist, and yet a third of global births are not registered. If your death is not registered then your wife (or husband) may have poisoned you and not been caught, but two thirds of global deaths are not registered. Registration of births and deaths also allows the generation of health statistics and matters for education, human rights, justice, equality, and security. Yet civil registration and vital statistics (CRVS in the jargon) is poor in most low and middle income countries and sometimes weak in high income countries, a meeting organised by the Lancet was told this week. The Lancet also published a collection of articles on the subject, a follow up to its 2007 collection. 

The global aim, said Carla Abouzahar of CAZ Consulting, is universal registration of all births and deaths, everybody being issued certificates, and the production of accurate, comprehensive, detailed, continuous, and timely statistics from the registration data. At the moment the world is a long way from this aim, but we should not let the best be the enemy of the good, said Alan Lopez from the University of Melbourne. It is possible, pointed out Richard Peto from Oxford University, to get reliable health statistics from a 1% representative sample of the population. Better good data from a few than poor quality data from many.

The essential data for health statistics, said Lopez, are for birth the number by year, the sex, and the age of the mother, and for death the number by year, age at death, and cause of death. Population data are also needed to provide a denominator. Some in the audience pointed out that other information like place of birth, nationality, and the names of parents was important for other reasons, and we were shown a copy of Barack Obama’s birth certificate to convince us of that.

High quality data on global births are not available, but there are good data on about one in six global deaths. Timely and reliable data are needed, said Lopez, because poor data lead to poor decisions and lost opportunities. (Unfortunately, pointed out Sarah Hawkes of UCL, the opposite of good data leading to good decisions and seized opportunities is not necessarily true.)

Despite many politicians and others recognising the importance of civil registration and vital statistics there has been little progress over decades, but, said Richard Horton of The Lancet, the time might now be right. Universal health coverage is becoming a major global objective, particularly following the poor response to the Ebola outbreak in West Africa, and knowing the “all” is clearly essential. Many of the Sustainable Development Goals will depend on good data, and most important of all may be how accountability is high on the global agenda. The achievement of objectives depends on monitoring, the democratic process, and action, and good data are essential for monitoring.

Although progress has been slow, there are now, said Abouzahar, regional initiatives in Africa, Asia/Pacific, and Eastern Mediterranean regions, and there are examples of countries making good progress—for example, Fiji and Kenya. There is also a global response involving many of the agencies of the United Nations, but it hasn’t risen to the top of the agenda—perhaps because so many bodies are involved. It’s an issue beyond health. Several mentioned that a motion from the UN General Assembly might prompt action.

The biggest challenge for health statistics is to get good data on the cause of death. This took a century in high income countries, and the data are still imperfect. Is the International Classification of Disease too complicated? Is it essential for physicians to record the cause of death when they are often not available or do it badly? The way forward may be to move verbal autopsies from being a research tool to routine use, and progress is being made with this aim. Information technology should be helpful, said Abouzahar, but we have to be aware of a proliferation of systems that don’t speak to each other.

David Phillips of the University of Washington has developed a method of scoring the quality and utility of vital statistics produced by individual countries, and it shows steady improvement in countries like Brazil, South Africa, and Mexico. But it also shows deterioration in countries like Ecuador, Sri Lanka, and surprisingly Switzerland. The problem with Switzerland is simply that it provides few data on cause of death, although the country has the data. Sometimes, said Lopez, an improvement in organisation can be enough to improve the quality of data because they are collected but simply not gathered together and reported.

Phillips and others have used the score of the quality of data to show in a complicated, controlled analysis that better data are associated with better health outcomes. Horton concluded that seven things are necessary to promote global improvement in civic registration and vital statistics: a business case; global leadership from a head of state or somebody similar; coordination of efforts; linking with civic society; money (the Bloomberg Foundation has just given $100 million); a UN General Assembly motion; and a scorecard to measure and report progress.

Richard Smith was the editor of The BMJ until 2004. He is now chair of the board of trustees of icddr,b [formerly International Centre for Diarrhoeal Disease Research, Bangladesh], and chair of the board of Patients Know Best. He is also a trustee of C3 Collaborating for Health.

Competing interest: RS spoke at the meeting, but realised that he knew less about the subject than anybody else in the room. He wasn’t paid and didn’t have any expenses, but he was given a free copy of the Lancet Collection.