Why your estimated due date might need a rethink

In the olden days, before smartphones and apps, I had quite the collection of cardboard pregnancy calculators. Around the time that I started researching estimated due dates, induction of labour, and the evidence relating to this (which was in about 2002), I began to seek them out with the enthusiasm of a niffler who had spotted a golden snitch.

Happily, they weren’t nearly as hard to find or catch.

At one time, every other organisation would give them out to midwives, obstetricians, and GPs. Conference stalls were overflowing with them and, if you mislaid yours, or got meconium on it, a colleague almost always had one they could spare. Usually tucked in the back of their paper diary, because we all had one of those back in the day as well.

The joys and tribulations of the whirly wheel

These cardboard calculators were designed so that, if you knew the date of a woman’s last menstrual period, you could work out her estimated due date. 

They often also contained handy reminders for midwives and doctors of when things should be offered, done or organised, according to the estimated due date. I have to say that the one with information about maternity benefits was so well used that it fell apart. That was truly useful information to have at my fingertips to share with women and families.

But an estimated due date, as you’ll know if you’ve read either of my books on induction of labour, is a guess with quite a wide timespan. Only five percent of babies (that’s one in twenty) are born on their due date.

Not quite as useful, then.

Pregnancy length varies

The thing is that, while it’s all very well calculating an average, variation is normal in nature.

Infact, in an important study on this topic, Jukic et al (2013) found that, even after excluding preterm births, the length of pregnancies varied by as much as 37 days. 

Factors which appeared to have an impact on the length of pregnancy in this study include maternal age (with older women having longer pregnancies), previous longer-than-average pregnancy and maternal weight at birth. 

“The length of human gestation varies considerably among healthy pregnancies, even when ovulation is accurately measured. This variability is greater than suggested by the clinical assignment of a single ‘due date’. The duration of previous pregnancies may provide a useful measure of a woman’s ‘natural’ length of pregnancy and may help in predicting an individual woman’s due date.” (Jukic et al 2013: 7).

Why we should personalise the due date

Other studies support the idea of personalising the estimated due date. For example, a paper by Ng and Steer (2016) showed that the prediction of a due date needs to take more factors into account than are currently considered.

Using information from over half a million pregnancies experienced by women who gave birth in London (which means it might not apply to everybody in the rest of the world, but it was a good-sized database including lots of different kinds of women), Ng and Steer (2016) looked at whether there was a relationship between the lengths of the women’s first and second pregnancies.

So their results are based upon the experiences of women who had more than one baby, but that doesn’t mean that their findings might not be useful more generally.

Indeed, there is so little research into this area that it’s really exciting to see studies which question the non-evidence-based practice of calculating and then relying on due dates based on universally used but arbitrary notions of when babies are due.

What did they find?

So what do their results show? Well, a couple of findings stand out:

1. Generally, shorter-than-average and longer-than-average pregnancies tend to recur.

So if a woman had a long pregnancy last time, the chances are she will again. Same if it was shorter.

This won’t be news to people who work with pregnant women, or probably to most women, in fact, but it’s always gratifying when research confirms what we already know from observation and experience.

This also confirms something that midwives in my own research on this area talked about a lot. That is, it’s more valuable to look at what is normal for the individual, than for the average of the population as a whole.

Regression to the mean

2. There is some evidence that the length of subsequent pregnancies in the same woman follows a mathematical phenomenon known as regression to the mean.

Don’t click away if you don’t know what that means … you don’t need to!

The fact that this data shows regression to the mean simply means that, if a woman’s first pregnancy is longer or shorter than average, the gestational length of the second pregnancy is likely to be closer to the ‘average’ (or mean) 280 days than the first pregnancy was.

So a woman who had a forty-two week pregnancy the first time is more likely to have one that is a bit shorter the second time around, though her pregnancy is still likely to last nearer forty-two weeks than thirty-eight. Same with a woman who had a shorter-than-average pregnancy and went into labour a week before her due date. The second time, she’s more likely to give birth (as long as we don’t interfere) somewhere between that time and her due date.

None of this means that we can predict exactly what will happen in any woman or pregnancy, of course. You might not be the average!

These data simply show us what is more or less likely, and the trends that we are looking at only appear if you look at lots of women.

If you are an individual woman reading this, then by all means do feel free to deviate from the population trend.

And if you are a birth professional, then by all means use this to further explain to women that our estimated and predictions really are only that. We might as well take up divination…

The Estimation Problem

But these things aren’t taken into account in reality.

In reality, we have apps as well as whirly wheels nowadays. And medical records systems, ultrasound machines and other technologies are also involved in calculating estimated due dates.

Which are problematic in themselves, especially when they are allowed to overrule human knowledge.

But pregnancy calculators all have the same problems in common.

They don’t all give the same date. Even when you give them the same starting point to work from.

And they certainly don’t consider the individual variation that I’ve been discussing.

The reliability problem

But let’s stay on the specific problem of their reliability.

I once tested whirly wheels by collecting about thirty of them, setting them all at the same mythical last menstrual period date and looking at the estimated due date they suggested. Later, some midwife friends around the world kindly did the same experiment for me on their hospital record systems.

When I compared the range of dates, I found that there was a whole week’s difference between the earliest and latest date given.

I later tried the same experiment with a series of free and paid phone apps, with the same result.

Even if you use the same LMP date, there is a few days variation in the estimated due date between the apps as well.

When you add that to the normal variation that we find in nature, it means that the due dates that we are using to estimate when babies will be born are nowhere near as reliable as some people think.

Why ultrasound isn’t the answer

Does that mean that ultrasound is the answer?

No. The variation exists there too.

Research which compared women’s estimated dates of birth as worked out from their LMP dates and ultrasound scans at a variety of gestational points with the date on which they actually gave birth found that scans were no more accurate than women’s own dates. 

The researchers concluded that, if the woman feels her own menstrual dates are reliable, there is no need for an early dating scan. 

This is good news for a couple of reasons. 

First, we really don’t know what the effects of such early scans are on developing babies, so it is good to know that these are unnecessary. 

But of just as much concern for me is the move that has occurred over the past few decades whereby technologically-derived knowledge has become seen as superior to women’s own knowledge, which also includes the knowledge that birth attendants can gain from women’s experience of being in their bodies.  

I don’t say that simply out of concern for women’s self-esteem (although of course that’s important).

I say it because I know that there are things we can know from women and their knowledge of their bodies that no machine will ever be able to tell us, and we need women to understand that this important kind of knowledge is valuable so that we don’t lose our connection with it. 

Women know this

Many women already know this. For instance, in a survey of 769 pregnant women planning a hospital birth in Australia, most of the women in the study understood that the expected date of birth (EDB) is imprecise.

But even more pertinently, the majority of women wished that the estimated timing of their birth could be given in less specific format than an EDB, with the majority of women preferring the idea of an estimated week or fortnight of birth.

There were some other interesting findings:

“Among the 769 women who responded to the question about expected timing of birth, 42% expected they would birth sometime before their due date, 16% sometime after their due date, 15% within a day or so of their due date, and 27% had no expectations. The expectations of nulliparous women were significantly different from those of multiparous women (χ² = 15.4, p = 0.002): nulliparous women were much more likely to expect to give birth before their due date.” (Todd et al 2016).

Expectations

And another insight that caught my attention from the same paper was this one:

“Women who completed the survey early in pregnancy (≤24 weeks) were more likely to have no expectations about their timing of birth; those between 25 and 36 weeks were more likely to expect birth sometime before their due date; and those at term (≥37 weeks) were more likely to expect birth after their due date.”  (Todd et al 2016).

I don’t know whether that’s of interest to you, but I tend to think that such snippets are worth sharing for those who like to consider the bigger picture.

But the main point I am trying to make here is about variation.

Back to variation

There are many examples of how variation is a fact of life when it comes to pregnancy, birth and estimating when these might happen.

It’s not rocket science to figure out why. We’re all different. As I wrote in In Your Own Time::

“In reality, bodies, babies and pregnancy lengths vary. In reality, we exhibit individual variation, and there is actually a wide span of time in which babies can be born and be healthy. In other words, normal is a range and not just one fixed point. One size hardly ever fits all.” (Wickham 2021).

One size didn’t fit all when I started researching this area more than two decades ago, and it still doesn’t now.

If you think that your due date might need a rethink, you’ve come to the right place.

I’ve written more on why we might be better off using a wider window, and on who is most likely to get an inaccurate due date. More on both of those in In Your Own Time as well.

Making induction decisions

The estimated due date affects when induction of labour is offered. Induction of labour is a very personal decision. It’s absolutely right for some women and families, and not right for others.

For even more info, visit my induction resources page, where you’ll find lots of articles like this one to inform, educate and inspire.

Post-term pregnancy and induction of labour resources

You might also like to read one of my books on induction: Inducing Labour: making informed decisions or In Your Own Time: How western medicine controls the start of labour and why this needs to stop.

Inducing Labour

In Your Own Time

And if you’re not sure which one you need to read, I have a blog post on that here.

Which induction book do I need?

 

---

---

---

Books to help you become informed and make the decisions that are right for you…

---

---