Maternal mortality reported trends in Afghanistan: Too good to be true? Analysis paper.

The Afghanistan Mortality Survey 2010 reports maternal mortality of 327 deaths per 100000 live births. An earlier estimate from 2002 of 1600 deaths/100000 live births suggests a rapid decline over 8 years a change that has been described as “too good to be true”. Our analysis indicates that there are indeed significant reasons to be sceptical of the figures generated by the survey. This is a critical concern given that maternal mortality is a key health development indicator and an exaggeration of progress in this area could affect funding decisions to the detriment of the women of Afghanistan and their newborns. This briefing explains why donors and policy makers need to be cautious in using the maternal mortality data presently available. Supposing the 2002 survey figure to have been approximately correct it would be difficult to explain the reported sharp decline in mortality when at the time of the 2010 survey two-thirds of deliveries still took place at home without skilled assistance. There are no precedents in other countries including those with greater stability than Afghanistan for such a major decline taking place on a time scale anything like as condensed as this data claims. If the decline were accurate Millennium Development Goal 5a to reduce the maternal mortality ratio by 75% from the 1990 level would have been achieved five years ahead of the target date the end of 2015.