Around one in three stillbirths occur before 28 weeks of pregnancy but are not officially recognised.
The burden of stillbirth has been underestimated by at least a third because of recommendations to report only stillbirths from 28 weeks' gestation in international comparisons, according to an observational study of 2.5 million babies in 19 European countries published in The Lancet.
The findings underscore the importance of accurate and consistent reporting of fetal deaths as early as 22 weeks so that the true burden of stillbirth can be understood and the impact on families acknowledged.
"There are major and serious gaps in our knowledge of the burden of stillbirth which will have significant unforeseen impacts on families," says Dr Lucy Smith, University of Leicester, UK, who led the research. "To a mother or father, a second trimester stillbirth is no less tragic than a stillbirth at 28 weeks of pregnancy or later. These parents also deserve recognition of their loss and accurate reporting of their child's death to improve care and policy." 
A Lancet stillbirth Series published in 2016 found that half of the 2.6 million stillbirths that occur worldwide every year (98% in developing countries) could be prevented, and estimated that, in high-income countries, for every 1000 births around 3.5 babies are born stillborn . However, the true burden could be substantially higher. WHO recommends that international comparisons use a cut-off of 28 weeks of gestational age. However, at a country level, WHO identifies 22 weeks of gestation as a threshold for identifying stillbirth, and there are also major differences between countries in their registration legislation (in part due to differences in the perception of viability of babies born early), as well as differing policies on termination of pregnancy being reported as stillbirths.
The authors used data from 19 European countries on pregnancy outcomes from 22 weeks of gestation between 2004 and 2015, to calculate overall rates of stillbirth and changes in rates between 2004 and 2015 by gestational age and country .
In 2015, more than 9300 babies were stillborn from 2.5 million births in Europe, and out of these a third were stillborn between 22 and less than 28 weeks of gestation, and would have been excluded from WHO's threshold for international comparison.
Between 2004 and 2015, the overall stillbirth rate between 24 and less than 28 weeks of gestation declined from nearly 10 to 7 per 10,000 total births, a reduction of 25% (figure 1B). This is similar to global figures of stillbirths over 28 weeks of gestation which fell 25.5% worldwide (from 247 to 184 per 10,000 births) and 24.5% in developed regions (from 45 to 34 per 10,000 births) between 2000 and 2015--suggesting consistent improvements in the reduction of stillbirths from 24 weeks of gestation over time .
Nevertheless, the variation across countries in stillbirths from 24 to less than 28 weeks in Europe ranged between 4 and 8 per 10,000 total births (when excluding terminations of pregnancy) implying that a large proportion of stillbirths are still preventable. These results are consistent with the variation in stillbirth rates at later gestation observed in many previous studies in high-income countries. The authors point out that the consistency in reporting of these earlier stillbirths suggests that these deaths should be routinely included in international comparisons to help inform clinical practice and policy.
In contrast, the overall rate of stillbirth between 22 and less than 24 weeks has remained unchanged since 2004, at around 5 per 10,000 births in 2015. The authors speculate that this is likely to be due to improvements in the reporting of deaths at these gestations. In 2015, rates of stillbirth (22 to <24 weeks) varied 6-fold between countries ranging from more than 2 per 10,000 births to 17 per 10,000 births (figure 1A).
According to Dr Smith: "Wide variation in the number of stillbirths occurring between 22 weeks and 24 weeks is likely to highlight differences in the collection of data across European countries rather than variation in underlying stillbirth risk. To ensure that the true magnitude and burden of stillbirth is understood, and to improve routine data collection for monitoring the outcomes and management of extremely preterm births from 22 weeks gestation, WHO's threshold for high-income country comparisons should be lowered." 
"Only through international studies, such as the Euro-Peristat project, that collect data with a standardised protocol, can reliable cross-country comparisons of stillbirth rates be made. This standardisation allows for consistent international reporting practices for terminations of pregnancy and agreed definitions for reporting of stillbirths and livebirths" says senior author Professor Jennifer Zeitlin, Inserm, Paris who coordinated the Euro-Peristat project. 
The authors note some limitations including that late terminations of pregnancy could not be excluded in a few countries, and the omission of some countries because of lack of comparable data for the study period, may have influenced the results--although sensitivity analyses supported the generalisability of the results as proportions of early stillbirths in these countries were very similar.
Writing in a linked Comment, Professor Marleen Temmerman from Aga Khan University in Kenya says: "In the Sustainable Development Goals, there is still no target for stillbirths, despite inclusion of a neonatal mortality target for the first time. This gap is a missed opportunity to count and drive change for stillbirths which are mostly preventable, but also a missed opportunity to count major impact that is coming from other investments for example in antenatal care and quality of care at childbirth."
She concludes: "Now is the time to count, and act on this burden, with good reporting from all countries above 28 weeks and all to collect data from 22 weeks. When countries meet data quality criteria, the 22-28 week stillbirth rates will be included in global reports. We also call for total births, livebirths, and stillbirths to be included in the denominators of key indices, such as maternal mortality, skilled attendance at birth, and Caesarean section rates, and for more innovation and investment in stillbirth data collection and use, including in surveys in low and middle income countries."
Another study published simultaneously in The Lancet finds that a reduced fetal movement (RFM) care package for pregnant women does not appear to reduce the risk of babies being born stillborn. The AFFIRM study is a randomised controlled trial of over 409,000 pregnancies from 33 hospitals across the UK and Ireland. However, the authors stress that their advice to pregnant women remains the same--if they notice any change in their babies' movements in the womb, they should contact their midwife or maternity unit immediately.
Peer-reviewed / Observational study / People