Last week a very lively training workshop full of the very best data geeks in Ghana and their Evidence for Action (E4A) collaborators set out to try to find the answer. How high - actually - is the maternal mortality ratio in Ghana? As the convener of the workshop I embarked on the task with great trepidation.
What is the real extent of maternal mortality in Ghana? And if we have asked every single household during the census - shouldn't we know the answer?Last week a very lively training workshop full of the very best data geeks in Ghana and their Evidence for Action (E4A) collaborators set out to try to find the answer. How high - actually - is the maternal mortality ratio in Ghana? As the convener of the workshop I embarked on the task with great trepidation.Participants at the workshop Even though I had one of the worlds leaders in demographic estimation methods Prof Allan Hill to lead - and the full weight of the Government Statistician Dr Philomena Nyarko behind the event - my thoughts strayed back to 2005 - to a meeting in New York - where the 'world' apparently agreed that measuring maternal mortality on a census is a good idea. Not only that - but that its recommended. And the rest of the world has listened! Since 2010 at least 15 countries have asked every household in their lands whether they have had a woman die in childbirth, pregnancy or shortly afterwards. Why the trepidation then? Especially as Ghana has included exactly the recommended wording in their census question - and are willing to use cutting edge techniques to process the data. The problem here is that the methods that we have to use to do the estimation - although tried and tested - are not conclusive. A number of interpretations are possible. Also - the method relies on adjusting the number of deaths - usually upwards - because many families - in Ghana as elsewhere in the world - do not like to report all of the deaths in their household. In fact they would - understandably - rather not talk about it. After the data geeks came up with a pretty good idea of the under-reporting factor in day 2 of the workshop - it became clear that the maternal mortality ratio estimated this way will be higher than expected. I was right to be anxious. But anxious feelings aside - I do feel that this was a very important workshop. Not only does it mark yet another groundbreaking meeting of a set of Ghanaian analysts from a mixed background of public health, statistics, academe and government (the MNH Working Group!)- who all care about the topic of maternal-newborn health - but this time there was a chance to look at data from every single household in the country.Despite the adjustments that are needed this will be an important analysis of maternal mortality - and subnational, urban versus rural and other interesting comparisons are possible. More fine-grained data investigation can also be done to find out what really determines maternal mortality in Ghana.Under the leadership of Dr Nyarko the forthcoming report from the workshop will lay out the new estimates. A debate will ensue about estimates from other sources and why they don't tally. But one thing is for sure - after asking everyone in the country about this issue-all of those involved with the workshop will ensure that the data are analysed carefully and that lessons are learned. And that a recommendation made ten years ago in New York City that measuring maternal mortality from a census was not in vain.