Although the global rate of decline in maternal mortality has accelerated since 2003, few countries are on track to reduce their maternal mortality ratio to the degree specified in the Millennium Development Goals, according to an analysis from the Global Burden of Disease Study.1 The researchers estimate that between 1990 and 2013, the annual number of maternal deaths worldwide fell from 376,000 to 293,000, and the maternal mortality ratio decreased from 283 to 209 deaths per 100,000 live births, equivalent to an average annual decline of 1.3%. The mean rate of decline was even greater—2.7% annually—during the final decade of the study period. However, from 1990 to 2013, only 16 countries, including seven in developing regions, attained the 5.5% average annual reduction needed to reach Millennium Development Goal 5, which set a target of a 75% reduction in the maternal mortality ratio (from its 1990 level) by 2015.

The new analysis used data from a range of sources—including vital registration data, sibling histories, census data, maternal mortality surveillance and verbal autopsies—to estimate levels of and trends in maternal mortality, identify the timing and key causes of maternal mortality, and make projections for 2030. Data were available for 180 of the 188 countries included in the analysis; for the others, the researchers relied on statistical models. Deaths were classified as maternal deaths if pregnancy was a direct or indirect cause, and if they occurred during pregnancy or within a year of delivery. In addition to small differences in data sources, the study differs from other recent global analyses in its inclusion of late maternal deaths (those from six weeks to one year after delivery) and in its treatment of deaths among women with HIV; on the basis of the limited data available from previous studies, the researchers assumed that 12% of HIV-related deaths during pregnancy and the first year postpartum can be attributed to maternal causes. To estimate maternal mortality by age, they created a variety of models using a subset of the data on nine variables associated with maternal mortality (e.g., age-specific fertility rates, proportion of deliveries occurring in a facility) and tested the models using the data that had been held back; the best performing model was used to make estimates for all available data. Because reliable data on births to 10–14-year-olds are scarce, maternal deaths in this age-group were not included in estimates of maternal mortality ratios, but they were included in overall counts of maternal mortality.

Additional models were used to estimate the proportion of maternal deaths that were attributable to each of nine key causes, and the proportion that occurred within four time windows: before the onset of labor; during labor or within 24 hours of delivery; 1 to 42 days after delivery; and six weeks to one year after delivery. Finally, the investigators estimated maternal mortality outcomes for 2030 in each country by extrapolating from the average rate of decline in the maternal mortality ratio from 2003 to 2013; if a country’s maternal mortality ratio had increased during that period, the authors assumed that it would remain constant between 2013 and 2030.

The researchers estimate that the annual number of maternal deaths worldwide fell from 376,000 to 293,000 between 1990 and 2013. Moreover, the maternal mortality ratio decreased from 283 to 209 deaths per 100,000 live births, which translates to an annual rate of decline of 1.3%; the average annual decline was greater between 2003 and 2013 (2.7%) than between 1990 and 2003 (0.3%). Decreases occurred in all age-groups, though risk varied widely with age; for example, the maternal mortality ratio among women aged 45–49 was 9.5 times that of women aged 20–24.

The average annual rate of decline in the maternal mortality ratio between 1990 and 2013 was smaller in developing countries (1.4%) than in developed ones (3.1%). Among developing regions, the decline was particularly striking in East Asia (8.9%) and, to a lesser extent, Andean Latin America (2.9%) and Southeast Asia (2.8%); in contrast, the ratio increased by an average of 2.7% annually in Southern Africa, despite an annual decline of 5.6% during the last 10 years of the study period.

Declines occurred in seven of the nine causes of maternal death. The biggest reduction was in the number of women dying each year from maternal hemorrhage, which fell from 71,000 to 44,000. However, the number of late maternal deaths and HIV-related deaths increased during the study period, and although the number of abortion-related deaths declined globally, it rose in Sub-Saharan Africa. HIV contributed to 0.4% of maternal deaths globally, compared with 1.5% of those in Sub-Saharan Africa.

About half of maternal deaths occurred during pregnancy (25%) or during or immediately after delivery (28%); a third occurred 1–42 days after delivery (36%) and the remainder (12%) were late maternal deaths.

Between 1990 and 2013, only 16 countries, including seven in developing regions, attained or exceeded the 5.5% annual decline in the maternal mortality ratio necessary to achieve Millennium Development Goal 5. Twenty-four other countries had that level of decline in either 1990–2003 or 2003–2013, but not over the course of the full study period. On the other hand, the maternal mortality ratio increased in 50 countries between 1990 and 2003, and in eight countries between 2003 and 2013.

If each country’s maternal mortality ratio continues to decline at its 2003–2013 pace (or remains at its current level, in the case of countries that had an increase during that period), then approximately 184,000 maternal deaths will occur worldwide in 2030. But even in this "fairly optimistic" scenario, the researchers note, 53 countries—nearly three-fourths of them in Sub-Saharan Africa—will still have maternal mortality ratios of more than 100 deaths per 100,000 births.

Limitations of the study include the likely misclassification of many deaths; the authors note that some deaths to pregnant women, such those related to injury, may have been incorrectly categorized as maternal deaths, while some true maternal deaths, such as those caused by unsafe abortion, may have been underreported. Other limitations include the substantial amount of uncertainty in the reported estimates, the sparse amount (or complete lack) of data for many countries, and the absence of a definitive approach for determining which deaths among pregnant women with HIV should be considered maternal deaths. Achieving further reductions in maternal mortality will require a multipronged approach, the researchers emphasize; key strategies will likely include promoting skilled birth attendance and delivery in facilities, discouraging early motherhood, reducing unsafe abortion and curbing malnutrition. In addition, the countries in which declines in maternal mortality have accelerated in recent years "should be carefully studied to provide qualitative insights into what has worked in different settings."—P. Doskoch

REFERENCE

1. Kassebaum NJ et al., Global, regional and national levels and causes of maternal mortality during 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013, Lancet, 2014, doi: 10.1016/S0140-6736(14)60696-6, accessed May 5, 2014.