Doyle and colleagues have reported the next largest data set to date, including live birth rates after warming 1,283 vitrified and subsequently warmed eggs between 2009 and 2015. This is one of the largest published data sets available. While not every embryo created from these frozen eggs was transferred, the authors estimated the live birth rates as if every viable blastocyst stage embryo was transferred. As a proxy for older patients who had fewer embryos, they used their success rates from their IVF outcomes – they had previously demonstrated that blastocyst stage embryos from frozen eggs had at least as high a pregnancy and live birth rate as blastocyst stage embryos from women of the same age using fresh, never frozen eggs. Their published outcomes involved the below calculations:
- Calculate the number of live births from frozen eggs from women at various ages.
- They still had embryos (blastocysts) remaining from the frozen eggs that were not transferred and refrozen, so they estimated how many of those frozen blastocyst embryos would result in a live birth based on their experience with blastocyst stage embryos frozen from women of the same age (at the time the eggs were retrieved). This estimate allowed them to approximate the efficiency per egg.
- Since this number looked unrealistically high for women at age 41-42, they substituted the efficinecy of live birth per fresh oocyte at this age and applied this number to the eggs that survived freezing/warming.
This study lacked extensive data on eggs frozen from women who are older than 37 (only 96 eggs). The 96 warmed eggs from women over age 37 yielded 5 babies. These numbers gave an unusually high efficiency for women over age 37 so the authors correctly adjusted their data to “predict” the efficiency using:
- Their survival rate for frozen eggs.
- The anticipated number of blastocysts that would develop from those eggs.
- Their live birth rates using blastocysts derived from fresh (non-frozen eggs) at each age group.
They then assumed a binomial “normal” distribution to yield the predicted probability per frozen oocyte at each age group. Our calculator uses the same assumption (a binomial distribution) to arrive at the probability of 1 or multiple children based on the number of eggs and the probability of a live birth per egg.
They arrived at the following estimates for live birth efficiency per egg:
Doyle et al 1,283 warmed eggs |
<30 | 8.2% |
<35 | 8.0% |
35-37 | 7.3% |
38-40 | 4.5% |
41-42 | 2.5% |