During our last pregnancy, I kept my expectations in check by remembering that we had had embryos which had not survived past the dates we were at. Margaret's pregnancy included a disappearing twin that reached about a 6 week size. Thus, even when we saw a heartbeat at a little less than 6 weeks, I figured we had a one in three chance of miscarriage, since of the three embryos that reached that size - the disappearing twin, Margaret, and Duncan - one hadn't survived. That kept the miscarriage from being too much of an emotional blow for me when it actually happened.
When our current pregnancy survived past that size with a healthy checkup at 8 weeks, it was tempting to assume that we were safe - or at least that our risks had dropped to the 1-2% level of an average pregnancy. That isn't really consistent with our history, though. Counting Margaret's disappearing twin and the blighted ovum that accompanied Duncan's pregnancy, along with the miscarriage, we've lost three of five previous embryos, far more than average. While pregnancy and the paleo diet have resolved Elizabeth's present endometriosis symptoms - apparently including the infertility - they haven't undone 38 years of effects on oocytes suspended in the dictyotene stage with the chromosome pairs partially separated. Endometriosis affects the microtubules that separate the chromosome pairs, and improper separation results in aneuploidy; aneuploidy was the likely cause of our previous embryos that didn't make it, so we're likely at increased risk of the aneuploidies that could survive past this stage as well.
The question is, how big is that risk? We've had three losses prior to 8 weeks. We have two healthy kids. We have no direct examples of losses after 8 weeks, but that's likely due to a small sample size; how do we estimate the actual probability of such a loss?
I think a reasonable assumption is that the expected proportion of miscarriages that occur after 8 weeks is the same for us as for the general population. The most credible figure I can find is that 80% of miscarriages occur in the first trimester. Considering some data on the distribution of miscarriages within the first trimester, I estimate that about 75% of miscarriages occur in the first 8 weeks. This percentage is roughly consistent with the percentage of trisomies that cause problems early enough that there's no chance of survival to birth - 17 of 23 - which I take to be independent verification that the 75% figure is reasonable.
So now we need to estimate what overall probability is associated with the remaining 25% of miscarriages that could still occur to us - even though that 25% has not previously hit us.
Well, our sample of completed pregnancies to date has been drawn from the population of early miscarriages in that 75% of all miscarriages, plus the population of successful pregnancies. In that sample, the miscarriages accounted for 3 of 3+2 pregnancies.
Our current pregnancy is drawn from the population of late miscarriages in the remaining 25% of all miscarriages, plus the population of successful pregnancies. My estimate for our miscarriage probability is thus (25/75)*3 of (25/75)*3 + 2 pregnancies, or 1 in 3.
It's actually a bit lower than that now, since we've had a couple of checkups since the 8 week point, and at the 12 week point, where 80% of miscarriages have passed, it will be 3 in 11, or about 27%. That's still pretty significant. Unfortunately, the odds don't really change much until we can verify or rule out the remaining trisomies with an amniocentesis at week 16. Until then, we just try to keep from getting too emotionally attached to this pregnancy.
Previous discussion of the effects of endometriosis on fertility: