Predicting success part 1: cumulative chances of success (or failure)

I’ve previously posted about my clinic’s success rates, depending on the quality of embryo used: http://3yearwait.blogspot.com/2014/04/get-out-your-calculator-its-time-for.html

Success statistics are amazing—in theory, they should tell you how many embryos to implant, and whether it’s worth your time and money to continue with further cycles.

Unfortunately, my experience has been that my clinic, at least, treats every cycle like it’s my first cycle—that is, they quote me success statistics based solely on the quality of the embryo and my age, not taking into account the obvious fact that each failure decreases the chances of success in the next round.  I also think they are wrong not to worry more about the lining thickness. 

Also, clinics often treat “pregnancy” as a success, but I think of success as a live birth of a child.  So those high pregnancy rates are not necessarily so wonderful, because the live birth rates are always lower.  (I’ve been pregnant 7 ¾* times, and I’ve had one live birth.  Pregnancy is not a success to me.)

A woman has created a blog that includes a “success rate calculator”:

http://www.smartfertilitychoices.com/ivf/

Now, an ideal success rates calculator would use all of your information – exact age, fertility diagnosis, embryo quality, number of embryos retrieved, whether the embryo is day 5, 6 or 7, whether the embryos are chromosomally normal, lining thickness and quality, body weight, results of previous transfers, etc etc etc. and give you a scientific chance of success based on all of your “data.”  This calculator asks for just a few of those data points (age range [35-39], diagnosis [unexplained], single or double embryo transfer [single], 3 or 5 day transfer [5], freezing method [vitrification]) to give you a chance of success per cycle.  She does not say exactly how she figures out her results, but here are hers for me:

	Chance of live birth
	fresh	frozen
Per cycle:	18.9%	23.1%
Per single embryo transfer:	25.2%	24.9%
Per year:	27.4%	37.4%

“Per cycle” means how likely are you to get pregnant if you start a cycle.  Per embryo transfer is how likely are you to get pregnant after you’ve actually transferred an embryo.  (She also gives a “per embryo” statistic separate from a per transfer statistic for people who are thinking about doing a two-embryo transfer.)  The last one is cumulative success rates for a woman who continues to try for one year.

This is somewhat interesting, but like I said it lacks input for other relevant data.  The most interesting outputs are the first two.  The third one (trying for a year) is nebulous.  Some women might get 6 frozen cycles in a year, others one or two.

Nonetheless, she paints a bleak picture for me—only a 25% chance of success per transfer, and my chances of having a child never go above even 40% if I try for a year.

Even without the robustness that would come with more data, I think she’s spot-on.  I’m screwed.

She also gives predictions for multiple cycle success rates.  For me, with a per embryo success rate of “25%” I have the following cumulative success rate (she does them in intervals of 2, so I’m optimistically using the 26% chance of success instead of rounding down to the 24% chance of success):

Cycle     Cumulative chance for a live birth

1             26%

2             45%

3             59%

4             70%

5             78%

6             84%

7             88%

8             91%

9             93%

10           95%

Wow, 95% looks amazing!  Again, though, that assumes 10 embryos to transfer.  Most women do not. 

Let’s frame this another way to demonstrate the real diminishing returns with more cycles:

Cycle     Chances cycle will work

1             26%

2             19%

3             14%

4             11%

5             8%

6             6%

7             4%

8             3%

9             2%

10           2%

That’s right.  You have a 26% shot with your first cycle.  If that one fails, there’s only a 19% chance the next one will work.  And so on until you only have a 2% chance of success.

That’s what I’ve been saying all along!  It’s not just simple math that you have the exact same chance with your first embryo as you do with your last embryo.  Presumably you used your better—and thus more likely to work—embryos first.  Thus, your chances are not the same at the end as they are at the beginning.  Even if your embryos are literally the exact same quality, your chances of success still decrease because whatever made the last embryo fail might doom the next one as well.

Her chart has cumulative odds from 2% per embryo (sucks to be at 2%) all the way up to 60% (young women, perfect eggs still can’t get above a 60% chance of success).  For those 60%ers, by cycle 5 99% of them will have a kid.  For the 2%, well there’s not a lot of hope. 

Here’s an article that recognizes, generally, that IVF has certain cumulative success rates depending on age:

“Among 6164 patients undergoing 14,248 cycles, the cumulative live-birth rate after 6 cycles was 72% … with the optimistic analysis and 51% … with the conservative analysis. Among patients who were younger than 35 years of age, the corresponding rates after six cycles were 86% … and 65% … Among patients who were 40 years of age or older, the corresponding rates were 42% … and 23% …. The cumulative live-birth rate decreased with increasing age….”

http://www.nejm.org/doi/full/10.1056/NEJMoa0803072?viewType=Print&viewClass=Print&_sm_au_=iVV3t8VMPNtsjRfP

As one article recognizes, “Contrary to the perception of many, IVF does not guarantee success; almost 38–49% of couples that start IVF will remain childless, even if they undergo six IVF cycles.”

http://en.journals.sid.ir/ViewPaper.aspx?ID=266706

That’s me.  Six cycles and no kid.  

The article recognizes many of the different factors that play into chances of success:

·        female age (younger is much better)

·        duration of subfertility (shorter is better)

·        type of subfertility (that a previous pregnancy or live birth substantially increases chances of success)

·        indication for IVF (male vs female, but it’s unclear which is worse)

·        basal follicle stimulating hormone (bFSH) (estimate of ovarian reserve, higher value is worse)

·        fertilization method (ICSI increases pregnancy rates)

·        number of oocytes (perfect number to retrieve is 15, and lower is worse)

·        number of embryos transferred (obviously more embryos transferred increases success rates)

·        embryo quality (obviously higher embryo quality is better)

It also cites 21 other articles that tried to model chances of success, defined as either pregnancy, ongoing pregnancy, or live birth.  Unfortunately, it does not propose a model, and the ones it says are good are not freely available.

Now, if my doctor’s off-the-cuff stats are to be believed (and I have complained already that they treat all cycles as easily likely to result in a pregnancy, which is just not true), here are my theoretical chances of pregnancy and live birth per embryo we have used:

	untested B2	untested B3	untested B3	normal B2	tested B2	tested B3
Approximate chances of pregnancy	51%	18%	18%	55%	55%	40%
Approximate chances of miscarriage	15%	15%	15%	8%	8%	8%
Approximate chances of live birth	43%	15%	15%	51%	51%	37%

If those statistics are true (which they are not) we have only a 6% chance that at this point all of those cycles would have ended in failure.  Which sounds low but that’s more than 1 in 20 women.

So, what are my chances of success for my last two embryos?  You got it.  NOT GOOD.

Here’s an article proposing a model for success, but it uses measurements that I either don’t know or don’t have:

http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3551786/

*Yea, this cycle did not work.  HCG of 1, which means it implanted and stopped growing.  This is the second time in a row that this has happened.  I will now refer to this as my ¼ pregnancy.  Who says you can’t be a little bit pregnant?