Are fewer mothers dying?

Almost 200,000 fewer women die each year from pregnancy-related complications than previously thought, because new survey methodology and better maternal mortality data mean more accurate mortality estimates, says a global study by the US-based University of Washington.



The most recent UN-funded assessment of worldwide maternal mortality estimated there were 535,900 deaths in 2005, while the new study put the number at 342,900 in 2008, after drawing on birth records, censuses, national surveys and interviews with next of kin and caretakers to determine causes of death.



Christopher Murray, one of the study's lead authors, told IRIN that new country data and better analysis had reduced errors. "Our study's data does not have information for 21 countries that account for 2.5 percent of live births, versus the [raw data used by the UN] that is missing information for 26 percent of live births."



The Institute for Health Metrics and Evaluation at the University of Washington created a database with three times more information than was used in the 2007 study of maternal mortality rates (MMR) commissioned by the World Health Organisation.



The data compiled by Murray's team – financed by The Bill & Melinda Gates Foundation – showed a 35-percent decline since 1980, while UN estimates showed that the number of mothers dying from pregnancy complications had changed little up to 2005.



More data



Monir Islam, head of the WHO programme, Making Pregnancies Safer, said he welcomed the findings, if they were accurate. "Most deliveries in the most affected countries take place in the home, so we do not even have data on those births," he told IRIN. "In conflict areas like Somalia, Afghanistan and the Democratic Republic of Congo, it is extremely complex to gather data."



Murray told IRIN that some maternal mortality data have been overlooked. "Even if births do not happen in health facilities, there are surveys that can estimate deaths [sibling history from Demographic Health Surveys funded by USAID], but people are sceptical of this methodology and it has not been widely used [to estimate MMR]."



The new study uses data from surveys about sibling deaths to estimate maternal mortality, but Murray said even in places where it was extremely difficult to gather data, there were ways that health epidemiologists were not using.



"There has been improvement in survey methods that have been around in other branches of science, such as remote sensing, agriculture and meteorology, but they have not been applied to global health."



WHO, the World Bank and the UN Population Fund will be releasing new global estimates of maternal mortality in May, followed by country estimates in October.



HIV



Murray told IRIN that not enough attention was paid to HIV as a cause of death - the study noted that more than 60,000 annual maternal deaths were related to lack of HIV medication.



"It is not emergency obstetric care that will make the biggest difference [in reducing the maternal mortality rate], but rather ARVs [antiretrovirals used to treat HIV/AIDS patients]."



Half the overall maternal deaths occurred in only six countries: India, Nigeria, Pakistan, Afghanistan, Ethiopia and the Democratic Republic of Congo.



Murray said the study showed some improvement in reducing maternal deaths but more needed to be done. "Scientific debate over statistics is always welcome ... What is equally important is to analyze what is working, and how to accelerate the reduction in maternal deaths."



pt/he