Due Date

Discuss

47 Responses to “Due Date”

  1. Anonymous says:

    I ranted about this before the birth of both of my daughters, and I count myself lucky to have a girlfriend who is actually as interested in this stuff as I am!

    The figure we were given was that only 2% of all children are born on their due date. What puzzled me was how such a small percentage could be the top of a Gauss bell. At least, I would expect the bell to centered on the due date, because what’s the meaning of an estimate if it’s not expected to be the most probable outcome (even by a very small margin)?

    One possibility we joked about is that of a Gauss bell centered around the due date, with a steep drop on the due date itself. In other words, the due date would give an indication as to when the baby will be born, with — say — a 10% chance on the day right before or after it, 8% on a two-day margin etc., but a much lower chance on the due date itself in the centre of the range. Which would be absurd, statistically, right? “She’ll be born AROUND January 1st but definitely not ON January 1st”.

    But if I imagine a Gauss curve with a true top that’s only 2% high, AND a limitation to about two weeks on the right, it has to be VERY wide at the left side, right? So I guess premature births of all kinds play a role in the statistics. Still, that 2% seems way too low to me.

    Rob (IANAS – I am not a statistician)

  2. Anonymous says:

    the distribution of human deliveries without intervention is a normal curve with mean 40 weeks and standard deviation of 1 week.
    you’re right that many people are induced or sectioned when they get beyond their due date in today’s environment – there is currently a national initiative to not electively deliver before 39 completed weeks but there are many reasons to deliver babies well before 37 weeks and a few reasons to deliver as early as 27 weeks – 13 standard deviations below the mean!!! so the normal curve is only for normal pregnancies

  3. TO Doula says:

    Others have commented on the silliness of trying to assess the scientific validity of a date that was based on fictional data to begin with (LMP), but the other variables that are important are parity (as mentioned, first babies tend to cook longer) and ethnicity. There is some evidence that fetuses of different ethnic origins have different average gestations.

    My theory about firsts often taking longer is related to why firsts tend to be longer labours, mamas may have issues with milk supply that they won’t have with subsequent babies, etc. The hormone regulation systems may benefit from practice, and certainly there’s a greater proliferation of hormone receptors on the uterus and an increase in milk ducts the next time around.

    So, to make the data you want to gather really meaningful, we’d need a few more pieces of information – parity, ethnicity, spontaneous onset of labour, health status, socioeconomic status. Maternal age may be a factor too, although I’m not sure if the increase in early births in older (first time especially) mothers is something about their bodies, or something about the reduced tolerance for anything that resembles risk (in both women and their care providers) when someone is over 40 and has spent $40,000 getting pregnant.

    I’d like to address the idea of risk with going postdates. Bryan3000000 said, “The number of things that can kill a baby being born late is astounding. The science is not on your side in this case.” I’d ask him to take a look at this document: http://www.kimjames.net/Data/Sites/3/understandingobstetricalrisk.pdf. My one semester course in statistics was long enough ago that I don’t feel qualified to say how good this is, but it does suggest that the risks of being “overdue” are not as great as many people fear. If you double the risk of a highly unlikely event, it’s still pretty damned unlikely. I’ve read Official Medical Documents that say pretty much the same thing. But that’s not the way it’s sold to tired, fed up pregnant women.

    (FWIW, my first came within 36 hours of her due date. My second was a full two weeks past it – and it was a good date, based on conception NOT LMP. Go figure.)

  4. roboton says:

    It was the damndest thing, I KNEW my daughter was going to born on May 13th that year. My wife kept saying “But it’s mother’s day”.

    She was born on May 13th.

  5. imag says:

    Seriously, that’s total BS. Unless, of course, someone has done the research, and it turned out that 280 days was actually the top of the bell curve.

    My assumption is that it would also vary by other obvious factors like temperature (weather related), etc..

    I have to believe that someone has done this.

  6. EH says:

    I thought that was what the “spin the dildo” game was for.

  7. Faustus says:

    Have you tried Pubmed? I just did a couple of searches but couldn’t find anything myself. It’s a hard topic to search for though as there’s a lot of relevant terminology.

  8. Anonymous says:

    Consider a due date like a politician’s promise. OK, that won’t work, no babies at all for us then. OK, more like baking a cake – it comes out when it’s done. The recipe might say 40mins on high, but everyone’s oven is different…

    Our 1st arrived 8 weeks early giving us a fair bit of a fright and almost 2 months of her stuck in hospital before she came home.

    6 months later and we have a happy smiley (but small) daughter making me happier every day.

  9. qatarperegrine says:

    I ranted about this too, while pregnant (this summer). I did find a few more stats and charts than you did. My recap is at http://www.zoecarlson.com/2010/07/28/due-date-probabilities/

  10. gabrielm says:

    Sup dawg, I heard you like statistics: http://bit.ly/eNV3NN

    But, seriously… I found this via a quick search for: “Human Gestation statistics”. It took me to this detailed post by spacefem, where she wondered the same thing and decided to start collecting data. Lots of good links on the subject.

  11. GeorgiKeith says:

    My guess is that it’s actually been pretty well documented, but that you can’t really talk about those numbers without discussing some uncomfortable topics like birth defects, and infant mortality; the numbers themselves are probably skewed early to accomodate the scheduled C-sections (like for breach babies). A quick google search for “Premature birth rates” shows that the statistics are known, and have increased in the last 30 years, since C-sections have gotten safer.

    If I were a doctor talking to a prospective parent, I would probably avoid talking about statistics too. There’s no reason for your doctor to say “well, X% of babies are born prematurely, but depending on how premature your baby is born, he/she has a f(x)% chance of survival.”

    In the end, I think the topic is far better discussed when you’re not pregnant… and I’m sure you’ll find people far more willing to discuss it then.

  12. Anonymous says:

    Uh-duh it is a bell curve! It is actually a due date window, that 95% of babies are born between 38-42 weeks, that being born between those gestational weeks increases survival as well and decreases stillbirth on the other end. Babies can be born outside of the window and be okay, but the farther you get from it the worse off they usually are with 44 weeks being the absolute max (they don’t like moms to go past 42 weeks but some do) but the lower limits being quite low.

  13. curleehair says:

    I think it’s a fascinating question. :D I had one on my due date and one 2 days early.

  14. jackie31337 says:

    I know I’m only a sample of 1, but here’s my experience:

    My due date was roughly estimated based on adding 40 weeks to the date of my last menstrual cycle. Throughout the pregnancy, it was adjusted based on the size/growth of my baby. In the end, she was born exactly 38 weeks (to the day) after she was conceived. It ended up being about 2 days after the most recent estimated due date.

  15. Am Elder says:

    There is research into this. A quick look at the EBSCO academic database reveals some recent and upcoming papers about the accuracy of due date predictions. So far, I’ve only found abstracts. I wont look too hard, since I’m not a doctor and I’m still a while away from taking a personal interest in pregnancy, but here are some of the titles of articles I found:

    - Are reported preterm birth rates reliable? An analysis of interhospital differences in the calculation of the weeks of gestation at delivery and preterm birth rate. BJOG: An International Journal of Obstetrics & Gynaecology; Feb2004.

    - The research implications of the selection of a gestational age estimation method. Paediatric & Perinatal Epidemiology; Sep2007.

    - Prediction of the date of delivery based on first trimester ultrasound measurements: An independent method from estimated date of conception. Journal of Maternal-Fetal & Neonatal Medicine; Jan2010.

    - Dating gestational age by last menstrual period, symphysis-fundal height, and ultrasound in urban Pakistan. International Journal of Gynecology & Obstetrics; Sep2010.

    - A comparison of first trimester measurements for prediction of delivery date. Journal of Maternal-Fetal & Neonatal Medicine; Jan2011.

    No disrespect meant by not citing the authors too. There are a lot of them and I figure they’re not important in a blog comment.

    I wonder if birthing professionals refrain from discussing due date accuracy because a) it can be exhausting to catch an amateur up on the science they’d need to have a full understanding of the issues involved, and b) they don’t want a nervous expecting parent to misunderstand the statistics to apply directly to their pregnancy experience.

    • bklynchris says:

      Dear God, thank you for this. I have a graduate degree in the answers to this and other related questions but the thought of finding the best links were about as appealing to me as reading my master’s thesis. Which, if I removed said thesis from under this desk’s leg, my laptop would fall off. Of course, if I removed said degree in embossed leather jacket from the opposite leg it would probably even out.

      But Mike, there are a variety of factors that are considered that in some ways make gestational age not necessarily linked to chronological fetal age. Most countries consider birthweight a far better morbidity and mortality indication of maternal and infant health than gestational age at birth. Gestational age is really a crap shoot based on a variety of factors, including but not limited to length of menstrual cycle, date of conception (always hard to prove unless mother was measuring fertility factors consistently previous to pregnancy), and sometimes date of intercourse (obviously least reliable). This is why when infants are born prematurely neonatologists consider other indicators such as Apgar scores, birthweight, lung health etc far more important than infant’s gestational age. The latter is really a number for OB’s who often recognize how it is often utterly arbitratry to the pregnancy at hand. So that’s why doctors “try not to think of it that way”.

      That said, thesis back under the desk leg with you.

      • Am Elder says:

        You’re welcome. :) Thank you for bringing a voice of knowledge. Oh what a tangled we weave when first we stick our noses into someone else’s area of expertise!

        When I typed “not important in a blog post”, what I meant was that the authors’ names wouldn’t mean much to most readers of the blog.

  16. nixiebunny says:

    I don’t really know when our children were due, due to lack of record keeping on our part. So there’s missing data as well.

    The spacefem distribution looks reasonable, since there’s a spike on the due date presumably caused by planned C-sections or induced labor. Busy working moms-to-be like to schedule these things well in advance.

    The preemie distribution has a nice asymptote to it.

    (Yes, I have designed Yagi antennas by “seat-of-my-pants” methods involving visual observation of the element spacing and length distribution, in case you were wondering.)

  17. warreno says:

    Prof. Brown – based solely on this post, I think I can see a little more clearly why you killed Pluto (so to speak).

    Asking a legitimate question about childbirth is, of course, sensible. Being surprised that medical professionals can’t give a decent answer is also sensible.

    Obsessing on the topic, analyzing it, and insisting to know for certain something that is – as you admit – unknowable seems a lot more like wanting to control the universe, or at least your little corner of it.

    That level of control is impossible. Bell-curve studies and planet declassification will not change that simple truth.

    Buy a bong, load the bowl, take a hit, and relax. That, or get a massage. Or take up meditation. You’ll go crazy otherwise, and you’ll take others to the nuthouse with you.

    • Avram / Moderator says:

      Unknowable? What’s unknowable about the statistical distribution of actual births around estimated due dates? As Mr Brown points out, it’s a trivial statistics problem for anyone who has access to the data.

  18. Anonymous says:

    Mike, I don’t know the answer, but the authors of this paper do.

    http://ije.oxfordjournals.org/content/33/1/107.full.pdf+html

    I would consider either digitizing the curve on Figure 1 and converting the cumulative frequency chart into the “bell curve” or writing the corresponding author of the study and requesting raw data.

  19. Anonymous says:

    This is the only data-driven study I’ve seen published. You may find it useful.

    http://www.ncbi.nlm.nih.gov/pubmed/2342739
    Obstet Gynecol. 1990 Jun;75(6):929-32.

    The length of uncomplicated human gestation.
    Mittendorf R, Williams MA, Berkey CS, Cotter PF.

    Department of Epidemiology, Harvard School of Public Health, Boston, Massachusetts.

    Comment in:

    Obstet Gynecol. 1990 Oct;76(4):732-4.

    Abstract
    By retrospective exclusion of gestations with known obstetric complications, maternal diseases, or unreliable menstrual histories, we found that uncomplicated, spontaneous-labor pregnancy in private-care white mothers is longer than Naegele’s rule predicts. For primiparas, the median duration of gestation from assumed ovulation to delivery was 274 days, significantly longer than the predicted 266 days (P = .0003). For multiparas, the median duration of pregnancy was 269 days, also significantly longer than the prediction (P = .019). Moreover, the median length of pregnancy in primiparas proved to be significantly longer than that for multiparas (P = .0032). Thus, this study suggests that when estimating a due date for private-care white patients, one should count back 3 months from the first day of the last menses, then add 15 days for primiparas or 10 days for multiparas, instead of using the common algorithm for Naegele’s rule.

  20. Glenn Fleishman says:

    I posted a little bit about this in passing a few weeks ago on BoingBoing in describing my hernia surgery (my surprise at how many bicycle parts were involved).

    Obstetricians, seemingly in particular, use a lot of oral knowledge. Across several obs we worked with across two births (for a variety of reasons), I found that a lot of what was planned for us was not rooted in clinical practice. It was often anecdotal, non-empirical, and even in opposition to mainstream research that was easy for me to find. Doctors of all stripes still struggle in the art/science split. Doctors want to believe that they aren’t machines, and using instinct and anecdote is one part of that. (There’s a lot to say that’s good about intuition and first-hand experience, but only when levied with provable, repeatable results.)

    The other reason for the variable delivery date is that the mother is asked to remember her last period before getting pregnant. I understand many women write this down. But the more appropriate date is conception. My wife and I know within a day or so the conception date of both our children; that’s never asked for nor used. So due date starts with an inaccurate number, and goes from there.

    • Loraan says:

      My understanding is that the dating is done from last menstrual period because, in general, pregnant women don’t know the date of their ovulation or conception. Menstruation, at least, has a visible external sign of its occurrence that most women recognize.

      I agree, however, that date from conception is a more relevant date, as the folicular phase is different for different women. A woman with a long or short cycle might be as much as a week later than “normal” in her ovulation, and this would throw her calculated gestational age off, resulting in a baby that seems to come “early” or “late.”

    • SamSam says:

      The more appropriate date to estimate from is not necessarily conception. The more appropriate date is fertilization, which is not the same as conception (if my conception we mean intercourse).

      The question is, which is a better estimator of fertilization, conception or LMP? Remember that sperm can stay alive for up to a week before fertilization. Ovulation, however, is generally a more precise time after LMP.

      Obviously there’s some variation in ovulation. I don’t know whether the variation is more or less than the variation between intercourse and fertilization. Either way, LMP should not be dismissed out of hand, as it may well be the better indicator.

  21. Anonymous says:

    For those that are pounding the drum on the impossibility of ever performing good analysis, what if we rephrased the question in these terms: “what is the CDF of the time span between reported LMP and the birth of the baby”? Look! No ambiguity! The beautiful thing about statistics is that it can make positive statements about uncertain events.

    When my first was born, I was able to use what statistics I could find to make strong statements to my team that there was a 25% chance that I would be out by this date, 50% chance by this date, and 90% by this date. That let them reason about what I should work on and what I should toss of my plate when.

    If we wanted a more accurate model, we could break it down by race, number of pregnancies, etc, and then figure out which are the most relevant external factors. This are the things math is good at. I was astounded that someone with a PhD and lots of post-doc work in the OB field wouldn’t have real statistics on-hand, but was disappointed too.

  22. Aknaton says:

    I do not remember where I heard that

    1) First children gestate for a week longer than later children, so it’s silly to use one number for them

    2) The standard length was determined by a study with a couple of hundred French women a century ago.

    The push to induce after the due date is outrageous. The PubMed articles make it clear that outcomes from induction are worse.

    • Loraan says:

      I have also heard that the average gestational age for first-time, white mothers is not 40 weeks, but 41.5 weeks. No source, although my girlfriend could probably find one.

  23. whatsupmonkey says:

    It’s worth remembering that the bell curve would vary over the course of the year. I briefly considered asking my mother (a midwife) to collect some data for me before realising that this month is December and I’ve been around midwives long enough to hear them talking about doctors inducing women early because they don’t want to get called in over Christmas.

  24. IronEdithKidd says:

    Another data point for you, Mike (first child, probably only child):

    Age of conception; 35. Ethnicity; caucasian. Initial estimated due date based on LMP; May 3. Estimate after 20 week ultrasound; May 5. Date of birth – no induction; May 8, 10:58 pm. My own calculation based on date of conception*; May 9.

    *We know date of conception due to my mate having been working and staying on the other side of the state. He was only home on the weekends and wasn’t able to make it home the weekend before conception. The weekend prior to that having been the LMP.

  25. Anonymous says:

    There’s been a shift in the bell curve, with length of gestation now being a full week shorter than it used to be, because of early inductions. http://californiawatch.org/health-and-welfare/early-elective-births-increase-so-do-health-risks-mother-child-7689

  26. Bryan3000000 says:

    As a guy who frequently has a problem with the way medicine is practiced, let me clue you in here – although you have a point about the accuracy of selecting the “due date”, you’re dead wrong about the bell curve at the higher end.

    The number of things that can kill a baby being born late is astounding. The science is not on your side in this case.

    Here’s the illustration. If someone isn’t breathing, are you going to wait to assist because there’s a bell curve in the rate at which people spontaneously begin breathing again on their own? They might do so without assistance, and if they fall significantly outside the norm, you’ll render assistance. The problem is, there’s an overlapping bell curve. It’s called the “point of death” bell curve. It applies to people who stop breathing and also to babies who fall at the long end of a normal period of gestation.

    When your wife and other ladies stare you down, please do yourself and the rest of humanity a favor and check your understanding. I had to eat my pride and learn this, and so should you. You happen to be picking on the one area where doctors generally have gotten it right, though there are still many doctors who err too much in this direction. But it’s easier to deal with a baby who could stay in a few more days than with one that’s already dead.

    There are plenty of areas where obstetricians and neonatologists are dead wrong in their practices. For example, good science has proven that constant maternal contact dramatically improves health and outcomes in premies. Yet doctors and nurses routinely keep mothers away and have nurses provide the bulk of the care. The doctors who started this practice are practically worshiped since they pioneered keeping the babies alive at all. But yet they were and are wrong in significant ways. Pick your fights more wisely.

    • bascule says:

      Bryan nice work with some insight into the science behind the question.

      You might want to reconsider your tone of voice though. You come across as pretty angry and condemning. I’m sure Mike will never swallow his pride and admit that he shouldn’t be asking questions.

      Mike was complaining about Doctors not telling him the info that you related in your post so there’s no point criticising him for not knowing the aforementioned info.

    • Anonymous says:

      Bryan3000000,

      You should learn the difference between people curiously asking questions and people who are claiming to be “dead right”. Mike Brown was asking questions and wondering about possible solutions. You come off like a jerk in your post and just make Mike’s intellectual curiosity all the more appealing rather than your frightened view of ever being wrong.

  27. Anonymous says:

    I couldn’t believe a study like that hadn’t been done. I found this on The Google: http://www.ncbi.nlm.nih.gov/pubmed/2342739?dopt=Abstract

  28. Anonymous says:

    This issue irritates me as well, for reasons others have posted. The mythological 28-day cycles is an AVERAGE. Many women have longer cycles.

    Despite SamSam’s assertion that ovulation is generally a precise time after LMP, this is not the case at all! I have been tracking my fertility signs to some degree for the past ten years. I have been pregnant several times, but during the times I was not pregnant, the date of my ovulation varied from 12 days after LMP all the way to 18-20 days! And I am a relatively normal fertile woman, with no hormonal imbalances or other factors affecting my fertility. There are many women with PCOS or other conditions affecting the length of their cycles, causing variations in the date of ovulation to be different than the standard assumption of 14 days after LMP.

    Most doctors do not even ask if their patients have been tracking their fertility signs, do not ask what the average length of their cycles is (a longer cycle is an accurate indicator that ovulation probably happens later than the assumed “cycle day 14″ average), and they do not believe their patients when they are told the date of fertilization is known. They will sometimes obstinately refuse to consider anything but their little cardboard wheel, even when the mother can provide a clear chart showing the rise in temperature indicating ovulation, or even the use of ovulation predictor tests.

    Women with later than average ovulation who track their fertility signs have taken to listing the date of their LMP as whatever day was 14 days before their known ovulation (i.e. lying about the date of their period) just to get the doctor to note an EDD that is not a week or two premature.

    I would be very interested in seeing a study of how accurate the EDD actually is, I just don’t have much confidence the scientific info would have a major effect on medical practice. I is known that a c-section rate over 15% increases risks for mothers and babies, and yet the rate is 30% or higher. It is known that episiotomy (intentional cutting of the genitals during second stage) used as a routine procedure increases damage and injury to mothers, and doesn’t improve outcomes for babies, but yet many doctors continue to do them as a matter of routine. I hope by the time my daughters are having children, the field of obstetrics will have finally caught up with their own research, but I’ll probably strongly advise them to see a midwife.

  29. CoquiELF says:

    I’ll throw in to the mix that both my brother and I were two weeks “late.” I told my OB I expected our daughter would also be two weeks late as, “it runs in the family.” As my due date passed, and OBs and medical insurance began freaking out more and more, pressure to be induced became very intense… then my water broke and my daughter was born… two weeks after my due date.

  30. Anonymous says:

    I was three weeks late, the doctors were going to induce labor, but according to my mom she went into labor unassisted because there was a full moon. Apparently more woman give birth when there is a full moon. Also for what it’s worth I was born on April 18 and only one of my nine siblings was born on their due date.

  31. qatarperegrine says:

    I think it’s popular somehow to think that moms make stupid decisions around their births, so I’d like to say two things in defense of pregnant women, having recently been one. :-)

    First, someone referenced the old “moms schedule C-sections these days because they’re too busy to wait” canard. It’s actually very rare for women to schedule a C-section just for kicks. (And most OBs won’t allow it.) The current rise in C-section rates has more to do with the rise in obesity, the rise in older moms, and the rise in multiple births — three situations that make C-sections much more likely.

    Also, someone said inductions lead to more complications. That’s true in a sense: a woman induced at 41 weeks 3 days, like I was, has more complications than a woman who goes into labor unassisted at 41 weeks 3 days. However, that’s not a fair comparison. If I hadn’t been induced, my alternative wasn’t going into labor unassisted the same day; my alternative was going into later unassisted at some unspecified point after that, OR needing induction at an even later point. If you compare it that way, postdate induction is less risky than people often say.

  32. qatarperegrine says:

    Oops, I forgot my main point. Sleep-deprived new parent here. :-)

    The latter point I made is why I so desperately wanted to see stats like this. When I was 41 weeks 3 days pregnant, what were the odds I’d go into labor naturally the next day? in two days? That information would have been very medically relevant for me. But, as others have said, OBs are too willing to say, “Oh, there’s no way to know.” I knew there was no way to predict when my daughter would be born, but there SHOULD have been a way to figure out the odds and make a rational decision based on them.

    As for the whole “LMPs aren’t accurate” thing, I got pregnant while living in a country (Qatar, if you can’t guess from my username) that does monthly ultrasounds. I understand early sonograms are quite accurate in pinning down gestational age, so it ought to be easy to pin down due dates better than we currently do.

  33. chgoliz says:

    I wonder if the *relative* lack of hard data might be partly due to the fact that this issue, by definition, is about women. When I think of the history of medical studies, I think of the many drugs and procedures that were tested only on men (heart problems, for example) and then the findings were automatically presumed to include women too, even though subsequent research has shown that is often problematical.

  34. MikeBrown says:

    First, I would like to simply say that you commenters are the best in the known universe. And I am thrilled to see that some of the data are now becoming available. As a data-seeker I am thrilled (sadly for me at the time, all of it is post-2005, which is when I cared the most!). But I am glad that the times are changing!

    A few specific thoughts, in reverse order:

    @bryan30000: No no no, sorry, don’t get me wrong. I was definitely not saying that I know what the answer is, I was just wildly speculating in the absence of any information and I was frustrated to be reduced to wild speculation!

    @byklchris: You certainly must be correct, that there are better methods than simply date. But, even so, there is an empirical distribution around the due date! That’s all I wanted to know, and had no way to find out. More detail would have been even better! [but see below....]

    @gabrielm, @qatarperigrine: you make my happier than any geeky once father-to-be deserves to be. thanks for sharing. if only they had done these back in the distant past (of 5 years ago….).

  35. Lobster says:

    You know what, Mike Brown? Your DAUGHTER isn’t a planet EITHER.

    How does it feel, huh? D:

  36. Anonymous says:

    Another question to pose is this: how useful are due date predictions anyway, given the old method of “days from last menstrual start day” assumes all women ovulate on day 14 of their cycle and have 28-day cycles. As most of the female readers know, cycle varies not just from woman to woman but from cycle to cycle for each woman! This renders the older method quite inaccurate, as it all depends on when a woman ovulates, not when her last period was.

    I like some of Elder’s article titles, specifically the one that attempts to predict due date based on ultrasound measurements. I shall have to delve further!

  37. singingdragon says:

    Something I’ve discovered is that doctors don’t actually have a whole lot of specific information at their fingertips beyond the questions they get asked all the time. No one has time to keep up with all the literature even within their specialty. Assuming you have a decent scientific background, you’re better off finding and reading some primary sources yourself than asking them an unusual question. You might even be better off reading secondary sources if that’s all you have access to.

Leave a Reply