The figures given in the league tables are the result of a new approach to measuring maternal mortality notoriously difficult to estimate because of both under reporting and misclassifying of deaths.
The new approach, developed by Johns Hopkins University, WHO and UNICEF, uses all available information including vital registration data, household surveys, sisterhood surveys (which ask women about the pregnancies and deaths of their sisters), and reproductive-age mortality surveys. To compensate for under reporting and misclassification, or for the non-availability of data in some countries, the new estimates also make use of a mathematical model based on two key independent variables the general fertility rate and the proportion of births attended by a trained person.
Despite data limitations, these revised estimates represent a major improvement on early efforts to quantify maternal mortality for all nations.
The country-by-country estimates given here differ, in some cases considerably, from previously estimated national statistics.
The new figures apply to the early 1990s. As a different estimating technique has been used, these figures should not be used for comparison with previous estimates in order to establish trends.