What, exactly, is amnioinfusion? It involves placing an intrauterine pressure catheter through the woman’s open cervix, past her baby, and into the amniotic fluid in her uterus. The same type of fluid that is used for intravenous administration is then delivered to the uterus. It’s been around since the mid 1970s (Weismiller, 1998), but never really taken off as a standard part of labour care. Amongst other things, it has been proposed as a way to improve outcomes when meconium stained liquor is present (Hofmeyr et al., 2014), and when there are decelerations presumed to be caused by umbilical cord compression (Hofmeyr et al., 2012).
A recent study from Israel has added to the body of research about using amnioinfusion when variable decelerations occur during labour (Cohen et al., 2025). Today’s post summarises what they found, and considers whether amnioinfusion should be more widely used or not.
First, the assumptions…
I want to start by logically stepping through what has to be true for this research, and their findings, to make sense. First, variable decelerations in labour would need to be (most of the time at least) due to partial or complete compression of the umbilical cord. The idea that cord compression triggers the baroreceptor reflex and this is the mechanism for variable decelerations, while head compression leads to early decelerations, and hypoxia triggers chemoreceptors and produces late decelerations doesn’t hold up as a physiologically plausible explanation however (Lear et al, 2018). So there’s immediately a fly in this particular bowl of soup.
Amnioinfusion is then assumed to work by reducing pressure on the umbilical cord and improving blood flow to and from the fetus. And finally, the disappearance of variable decelerations is assumed to indicate that cord compression has been resolved, thereby improving fetal oxygen levels. With these assumptions made visible, let me push on to see what this research team did.
What did they do?
This was a retrospective pre-and-post intervention study, not a randomised controlled trial. They studied a group of 131 women who had amnioinfusion for recurrent variable decelerations that persisted despite position change, and where the CTG trace was interpretable. It is important to note that 84 women, almost 40% of those potentially eligible, were excluded due to the quality of the CTG. This is worth keeping in mind when thinking about how to apply these findings in the messy world of clinical practice, where CTGs are often messy too.
The main outcome in this study was not the presence of variable decelerations but the total deceleration area: a measure of both the time the heart rate was below baseline and how low the heart rate was. The total deceleration area seems to have been calculated by obstetricians who were not aware of the purpose of the study or the outcomes of the birth, well after the birth had occurred (this is a good idea as it reduces ascertainment bias). Mode of birth and neonatal outcomes were recorded – but there was no comparison with a group of babies where amnioinfusion was not used.
What did they find?
The total deceleration area 30 minutes before amnioinfusion was 8430, and 30 minutes after amnio infusion was 3590, a statistically significant reduction. Amnioinfusion appeared more effective when the amniotic fluid index (measured by ultrasound on admission) was normal rather than low, and when umbilical cord entanglement was present at the time of birth.
For the women enrolled in this study, the caesarean section rate was 34%, and the instrumental birth rate was 22%, leaving a low vaginal birth rate of 44%. There were no perinatal deaths or cases of hypoxic ischaemic encephalopathy. Only 2% of babies had an Apgar of under seven at five minutes and 8% had cord blood acidosis at birth. There is no way to determine whether amnioinfusion altered the mode of birth or outcomes for the baby, as the study design did not allow for this.
What does this mean?
This study is slightly confusing (starting with women with variable decelerations but then not measuring these) and the design doesn’t provide answers that are particularly useful in clinical practice. No one knows whether a reduction in total deceleration area after amnioinfusion means that fetal oxygen levels are improved, or whether the fetal ability to compensate for low oxygen levels (by slowing heart rate and reducing myocardial oxygen requirements) is perhaps somehow being impaired instead. The authors jump ahead of the evidence claiming that their findings “suggest a potentially beneficial quantitative effect in reducing both neonatal acidaemia and neonatal morbidity”. I don’t believe that anyone should be adopting amnioinfusion on the basis of the current evidence.
Having skills in reading and interpreting research is hugely important. Rushing off and offering amnioinfusion on the basis of a paper like this, simply because the authors write something overly optimistic in their discussion would be very much the wrong thing to do. That isn’t what evidence-based care looks like. This is the reason I specifically included a module about how to interpret research in the Fetal Monitoring Academy, and why I set through my own analytical choices and justify them in the Academy lessons. If you want to improve your research reading skills, check out the Academy while enrolments are open!
You’ve read the blog posts, but still want more? Wish you could find details summaries of all the evidence in the one place? Want to connect with a growing tribe of people working to solve the fetal monitoring problem?
Now you can!
The Fetal Monitoring Academy is a HUGE resource library covering just about every aspect of fetal monitoring research, with expert analysis and critique so it is easy to understand. New content is added regularly in response to requests from Academy members – so you will find exactly what you are looking for. You also get:
- Ready access to leading expert Dr Kirsten Small to have your questions answered
- Monthly live Meetups to discuss aspects of fetal monitoring practice, with a view to promoting and supporting practice change
- A secure online community space where like-minded people hang out. It’s place to be supported and to support others as we work together to stop the nonsense
- Downloadable resources
- Certificates you can use to claim professional development hours
If you are a maternity professional, doula, educator, researcher, or simply a birth nerd who wants to know more – the Fetal Monitoring Academy is perfect for you.
Enrolments are open for a short time only each February and August. Don’t miss this window of opportunity! Enrol today.
Enrolments are open now!
References
Cohen, G., Bar Noy-Traub, N., Schreiber, H., et al. (2025). Is amnioinfusion for intrapartum variable decelerations effective? Evaluation of the total deceleration area: A retrospective cohort study. International Journal of Gynecology & Obstetrics, 00, 1-9. doi:10.1002/ijgo.70239
Hofmeyr, G.J., & Lawrie, T.A. (2012). Amnioinfusion for potential or suspected umbilical cord compression in labour. Cochrane Database of Systematic Reviews, 1, CD000013. DOI: 10.1002/14651858.CD000013.pub2.
Hofmeyr, G.J., Xu, H., & Eke, A.C. (2014). Amnioinfusion for meconium‐stained liquor in labour. Cochrane Database of Systematic Reviews, 1, CD000014. https://doi.org/10.1002/14651858.CD000014.pub4
Lear, C. A., Westgate, J., Ugwumadu, A., Nijhuis, J. G., Stone, P. R., Georgieva, A., Ikeda, T., Wassink, G., Bennet, L., & Gunn, A. J. (2018, Dec). Understanding fetal heart rate patterns that may predict antenatal and intrapartum neural injury. Seminars in Pediatric Neurology, 28, 3-16. https://doi.org/10.1016/j.spen.2018.05.002
Weismiller, D. (1998). Transcervical amnioinfusion. American Family Physician, 57(3), 504-510. https://www.aafp.org/pubs/afp/issues/1998/0201/p504.html
Categories: CTG, EFM, New research
Tags: amnioinfusion
Birth Small Talk 