Podcast: Interventions for improving outcomes in pregnancies that follow stillbirth

The Cochrane Pregnancy and Childbirth Group has produced several hundred reviews on the effects of interventions for women who are pregnant, giving birth or with a newborn baby. They also assess the evidence on how to help after pregnancy loss, and one of their new reviews from December 2018 looks at care during pregnancies following a stillbirth. Lead author, Aleena Wojcieszek from the Mater Research Institute in the University of Queensland in Brisbane, Australia, tells us what they found.

"Women who have had a stillborn baby are up to five times more likely to have a stillbirth in their subsequent pregnancies. They are also at higher risk of other complications, such as preterm birth and growth problems for the baby. On top of this, women and their families often experience high anxiety in pregnancies after stillbirth, which can be particularly intense at significant pregnancy milestones and in the lead up to, and during, antenatal care visits. However, although these women and their families clearly have increased medical and psychosocial needs, there is currently little evidence to guide clinical practice and provide optimal care.

We did this broad systematic review to bring the existing evidence together, aiming to capture any interventions or models of care that might be provided before and during pregnancies following stillbirth. These could include medical interventions, as well as psychosocial support targeting anxiety, depression or complicated grief. We focussed on interventions for mothers or their partners whose previous stillbirth occurred 20 weeks or more into the pregnancy.

The lack of research is highlighted by the fact that we were able to include just 10 trials, with data from only 222 women and their babies. All these trials assessed drug therapies and presented us with nine different comparisons, including looking at the effects of low dose aspirin, low molecular weight heparin, third‐party leukocyte immunisation, intravenous immunoglobulin G, and progestogen.

With so few participants, such a small number of trials, and so many comparisons, this relatively small amount of evidence means that it is uncertain whether any of the interventions had an effect on most of the outcomes we looked at. Although there was some suggestion that low dose aspirin and third‐party leukocyte immunisation may increase the birthweight of babies, even these findings were unstable due to the very small sample sizes.

One of our greatest challenges was when assessing the articles for eligibility. There was a lot of inconsistency in the terminology used to describe pregnancy loss, and reports often did not define words like ‘miscarriage’, ‘abortion’, or ‘stillbirth’ by gestational age. We often had to contact the original researchers to determine if a trial was eligible, and we also found that in many cases women whose previous pregnancy loss occurred 20 or more weeks into their pregnancy were recruited alongside women who had earlier pregnancy losses. This lead to the further complication of needing the researchers to split up their data, so that we could use only data from the former women in our analyses. We are very grateful to the researchers who did this and provided data, but it’s also important to note that not all trials in the review necessarily targeted the pathophysiology of stillbirth specifically, and we had to leave out some potentially eligible trials because researchers did not respond to our queries or for other reasons.

In summary, we were not able to capture sufficient evidence to inform clinical practice for care in subsequent pregnancies following stillbirth. Carefully-designed trials are urgently needed, and these trials should specifically address the prevention of recurrent stillbirth, and be large enough to detect differences in rare, but vitally important outcomes such as stillbirth and death of the newborn baby. We would also like to see the prioritisation of trials of psychosocial support interventions, which are currently lacking from our review."