President Bush
is poised to veto
another stem-cell bill any day now. I'm waiting with baited breath. During the interim however, I've been looking into so-called "
embryonic adoption" (the practice of taking frozen embryos, which would otherwise be discarded or go to scientific research, and turning them into babies). These little frozen guys have been called "
souls on ice", "frosties", "snowflakes" and other equally adorable names, and they are the Bush administration's answer to the
IVF dilemma. With the recent chatter about the implications of
the newest stem-cell advance (in particular,
a comment made by a fellow blogger), I got to wondering, what's the real chance of an embryo becoming a baby once adopted?
The success rate of
in vitro fertilization is a
complicated statistic to measure. The laundry-list of compounding variables is extensive: the age of the mother; the father's fertility; the mother's general health; the complexity of the infertility issue; the general practices of the clinic, to name a few. Moreover, individual clinics advertise numbers to entice "buyers" (even though they must report their
actual success rates to the CDC annually), and they don't always reflect easily understand statistics*. Finally, clinic-reported success rates are usually expressed in a population level manner (summed, averaged, totals), whereas to understand how many adopted embryos will become babies, we need only the data for the "frozen donor embryos".

The best source then, for this type of data is the aforementioned annual CDC report. The
most recent report available is from 2004, so the data may be a little out of date (in general success rates have gone up annually since 1978). Unfortunately, the report doesn't detail frozen donor success rates very well, so the following
back-of-the-envelope calculations are necessarily derived from the frozen nondonor embryo data (there is only a slight difference in frozen nondonor and donor data, with the success rate for donor embryos being marginally higher). The total number of live-births from nondonor frozen embryos in 2004 was 4,658. However, the live-birth data isn't the number of infants born, so a bit of math is necessary to find the actual birth rate. Using the Figure 38 (pg. 51 of the report), then, and counting multiple-non-twin-births as quadruplets†, we find that there were about 5,981 actual infants born from frozen nondonor embryos in 2004. In other words, females implanting their own thawed embryos had about 5,981 babies that year.
Next, in order to determine the number of embryos transferred per actual birth, we need to turn to the national summary data (pg. 81). In particular we are interested in the section called "Frozen Embryos from Nondonor Eggs".

In total, 16,296 transfers of this type were preformed on women under age 42. However, "transfer" here doesn't mean number-of-embryos-transferred but rather, the number of transfer-events recorded. So, in order to figure out how many actual embryos were transferred, we have to find the weighted average of "Average number of embryos transferred" (~2.58) and multiple it by the total number of transfers (16,296). As it turns out, approximately 41,982 frozen nondonor embryos were transferred in 2004, from which 5,981 births were recorded. A 14.4% success rate. Or expressed another way, for every infant born from a thawed embryo, six embryos didn't make it.
.jpg)
The reasons for this low success rate are complicated. Frozen embryos aren't as hardy as fresh embryos, in general (graph left). Moreover, the best embryos have usually been selected and used during the first cycle of implantation, so the frozen embryos are the "worst of the crop". Also, the embryos, though seemingly viable, often don't implant right or spontaneously abort sometime during the pregnancy for unknown reasons. In addition, if multiple embryos are successfully implanted, parents will frequently elect to selectively terminate some very early in the pregnancy (multiple births from IVF can cause a host of complications best avoided by early termination). Finally, all of the general complications associated with IVF can come into play.
The embryonic stem cell debate revolves around the intrinsic value of the human embryo. Even the most fervent supporters of the research view human embryos as a special entity, when say, compared with mouse embryos. Their inherent
potential to become human beings makes them unique in the eyes of all. President Bush went so far as to liken
the destruction of this potential to murder when he vetoed
the first Stem Cell Research Enhancement Act back 2006: "[T]his bill would support the taking of innocent human life in the hope of finding medical benefits for others." (
1)
However, what the above approximate calculations reveal, is that even in a best-case scenario, where all 500,000 frozen embryos are adopted, about 428,000 have
no potential, they just won't make it. In truth, about 97% of the "frosties" will never be adopted anyway and simply expire in the freezer (no use to anyone). Not the most pleasant of thoughts, but a reality that the Bush administration and many others have simply ignored.
*The "success rate" of clinics is a somewhat odd static often measured in live-births-per-cycle or live-births-per-transfer, where a cycle is egg extraction, fertilization, transfer (of multiple embryos), and pregnancy and a live-birth is mostly what is sounds like, except that it is counted as one live-birth, no matter how many babies are actually born/survive past birth.
†I count all multiple-non-twin-births as quadruplets in an attempt to account for the discrepancy between frozen donor and frozen nondonor rates -- somewhat arbitrary, but all calculations are offered only as approximates.