How the argument for intrapartum CTG monitoring is set up
There are two consecutive logic steps used to generate the argument for intrapartum CTG monitoring seen in obstetric guidelines, education programs, and other documents. The first step is to establish that population group A is at higher risk for a poor outcome than population group B. So for example, it might be argued that women who have a small fetus are at higher risk of stillbirth. The second step is to argue that using intervention X, in this case intrapartum CTG monitoring, reduces the risk of this outcome. This risk management approach is widespread in maternity care and in healthcare in general.
One of the reasons that ineffective or under-assessed interventions are introduced and then persist in clinical practice is because step one happens and a risk identification process is introduced into practice. When there is knowledge that an individual has a higher risk for a poor outcome, this generates (appropriately) a desire to do something to prevent the poor outcome from happening. A problem arises when our concern for the wellbeing of the woman and her baby see us reaching out and applying interventions that we think should work, but where there is evidence of no benefit from this approach.
Changing the focus to risk
In previous blog posts I have explored the evidence for step two and demonstrated that it is not as sound as we imagine it to be. (You can find posts about this here, here, here, and here). In today’s post I want to offer a glimpse at what seems to be going on in step one – the risk assessment process.
At risk for what?
This is a very important question to ask. Often the use of the term “high risk” in maternity care is vague and nebulous, coloured with fear and mystery. Intrapartum CTG monitoring was created to prevent some specific poor outcomes and not others. A woman might be at high risk for poor mental health or deep vein thrombosis after the birth of her baby. Using a CTG is not going to help with these risks, so risk assessment should focus on the sorts of things that CTG monitoring was meant to help prevent.
Intrapartum CTG monitoring was intended to prevent death (during labour or after) or damage (mostly to the developing brain) to the baby due to low oxygen levels occurring during labour. These outcomes are:
- Intrapartum stillbirth due to intrapartum hypoxia
- Neonatal death due to intrapartum hypoxia
- Cerebral palsy due to intrapartum hypoxia
- Hypoxic ischaemic encephalopathy due to an intrapartum cause
CTG monitoring in labour can’t prevent:
- Stillbirth before the start of labour (antepartum stillbirth) because the event happens before the monitoring would occur,
- Neonatal death due to causes that happened before the start of labour, or were not related to low oxygen levels, and
- Damage to the baby that happened before the start of labour, or was not related to low oxygen levels.
When examining potential risk factors to include as indications for intrapartum monitoring, we ideally should only include those that increase the risk of outcomes that might be modified by intrapartum monitoring.
Establishing a baseline
In order to decide what is higher risk, we need to establish the baseline risk. This is tricky. Risk is always context dependent. For instance, the risks for a woman giving birth in the USA is different to the risks for a woman giving birth in Britain. Finding a single universal truth is simply impossible. Good fine grained data are not easy to come by and I have not yet found a report that clearly distinguishes outcomes by timing of injury and cause. The best I have managed to find comes from the detailed reports of the Birthplace study (Hollowell, et al., 2011) conducted in the UK, and the Birthplace in Australia study (Homer, et al., 2019). I present the figures here with the caveat that they may or may not be relevant to the context that you are in.
Among women with a singleton fetus of at least 37 weeks gestation, the following outcomes were reported in these studies
- Intrapartum stillbirth – 2.6 to 3.2 per 10,000 births
- Hypoxic ischaemic encephalopathy (either a clinical diagnosis or signs to imply this diagnosis) – 17.9 per 10,000 births (UK data only)
- Neonatal death in the first seven days of life – 2.7 – 3.2 per 10,000 births
Only short term outcomes were examined in the two Birthplace studies. Cerebral palsy is said to occur in 2-3 per 1000 live born infants, with between 3 and 50% of these believed to relate to events occurring during labour (Philpot, Greenspan, & Aghai, 2020). If we assume a rate of 20%, then 6 babies per 10,000 will develop cerebral palsy due to intrapartum hypoxic events.
Adding these up gets us to a combined rate of 29.2 – 30.3 per 10,000 births, or 0.3%. This is still an over estimate as some babies will have a diagnosis of encephalopathy and go on to die, or develop cerebral palsy and adding them up in this way counts them twice. These numbers include babies who died or where injured from causes like trauma or infection, rather than low oxygen levels.
A sign that risk assessment isn’t working
So, we have a set of outcomes that occur for fewer than 0.3% of women giving birth to a single baby after 37 weeks of gestation. An ideal risk assessment process would generate two groups of women. A large group with a much lower risk than 0.3%, and a small group with a much higher risk than 0.3%. But that’s not what we see in practice.
In the data I gathered for my doctoral research, 90% of women were considered to be high risk and were therefore having CTG monitoring during labour. This rate of CTG use was similar with statistics for all of the state, and when I speak with midwives from the UK and Ireland, they report similar rates. How can it be logical that 90% of women giving birth are at risk for an outcome that impacts 0.3% of women? This seems to me as a strong sign that when it comes to intrapartum CTG monitoring, our risk assessment processes aren’t working they way they should.
Hollowell, J., Puddicombe, D., Rowe, R., Linsell, L., Hardy, P., Stewart, M., Redshaw, M., Newburn, M., McCourt, C., Sandall, J., Macfarlane, A., Silverton, L., Brocklehurst, P., & Birthplace in England Collaborative Group. (2011). The Birthplace national prospective cohort study: perinatal and maternal outcomes by planned place of birth Birthplace in England research programme. Final report part 4. http://openaccess.city.ac.uk/3650/1/Birthplace_Clinical_Report_SDO_FR4_08-1604-140_V03.pdf
Homer, C. S. E., Cheah, S. L., Rossiter, C., Dahlen, H. G., Ellwood, D., Foureur, M. J., Forster, D. A., McLachlan, H. L., Oats, J. J. N., Sibbritt, D., Thornton, C., & Scarf, V. L. (2019, Oct 29). Maternal and perinatal outcomes by planned place of birth in Australia 2000 – 2012: a linked population data study. BMJ Open, 9(10), e029192. https://doi.org/10.1136/bmjopen-2019-029192
Philpot, P., Greenspan, J., & Aghai, Z. H. (2020). Problems during delivery as an etiology of cerebral palsy in full-term infants. In Cerebral Palsy (pp. 67-76). https://doi.org/10.1007/978-3-319-74558-9_6
Categories: Basics, CTG, EFM, Perinatal brain injury, Perinatal mortality, Stillbirth
Thank you, Kirsten. I absolutely LOVED your interview on the Birth Rebellion podcast. As a birthing woman, I’ve observed that fetal heart rate monitoring in labour, and its implications, is not often discussed in birth education (save advice on the different forms of monitoring, and broad guidance that CTG for low risk labours is generally not recommended). It’s also an area I suspect the average ‘layperson’ is somewhat reluctant to self educate around, given both the emotive and complex nature of the issue.
For me, I believe continuous monitoring was responsible for very enthusiastic coached pushing during the second stage (ie ‘baby’s not too happy you really need to get them out now’), resulting in what I experienced as a traumatising level of pain I couldn’t control. CTG was recommended after decels where heard with the Doppler. It is quite unbelievable to learn that no research has been undertaken to explore whether progressing to CTG is effective when something abnormal is heard during intermittent auscultation!
Following on from my experience, I’d love to ask a few questions. To go back to first principles, I’m wondering, couldn’t we only say that intermittent auscultation is effective if we compare it with no monitoring at all?
Second, if we aren’t really sure what normal and abnormal heart rates look like on a CTG (because we don’t well understand physiology) how do we know what normal and abnormal look like with intermittent auscultation? Or is this a moot point because in the hospital system, we’ll always progress to CTG if something understood as ‘abnormal’ is detected?
What can we learn from home birth, where we may not have the option to progress to continuous monitoring? Can IA from an experienced midwife discern between a baby that’s compensating but is fine, a baby that needs to come out, and a baby that REALLY needs to come out, better than chance alone?
Does continuous monitoring and IA identify a happy baby and a very unhappy baby at the extremes, but continuous monitoring is not good at interpreting the scenarios in between (leading to a lot of unnecessary intervention)? Therefore, monitoring is very necessary but we need to be aware that continuous monitoring is associated with unnecessary intervention?
It is chilling to hear you confirm the variability between and within practitioner assessments of fetal heart rate, as this is something I suspected after reflecting on my own birth experience. Yet another way in which we play roulette with our birth experience.
Apologies for the essay and all the questions but I’m so glad you’re researching so passionately in this area, and that it’s finally making its way into mainstream birth discourse!!
Hi Georgie, Thank you for taking the time to read and consider the issues. In answer to your questions 1. There has never been research comparing no fetal heart rate monitoring with intermittent auscultation, so we don’t know whether intermittent auscultation is better than nothing. 2. The same physiological principles apply whether you hear the change on intermittent auscultation or see it on the CTG. It is my belief (not evidence) that we will never get fetal heart rate monitoring to “work” as it doesn’t actually measure what we need it to. Instead, I think we need to focus on identifying other, actually effective ways to improve outcomes that can replace fetal heart rate monitoring or make it irrelevant.
Fascinating. Fundamental questions around fetal heart rate monitoring in and of itself, let alone continuous monitoring. Thanks.