Predicting IVF Success Rates

In Vitro Fertilization (IVF) is a way to address fertility problems. In IVF, a woman's egg is removed from her and combined with donor sperm (possibly a known donor) and combined in a lab. "If a healthy embryo develops, it is reinserted into the woman’s uterus. In the best-case scenario, this embryo attaches to the uterine wall and a normal pregnancy ensues." The trouble is that it often doesn't work (the success rate is about 27%), and each attempt costs $7,000 to $15,000.

A new system much more accurately judges the chances of success of a second try than previous methods (in the form of a simple predicted percentage likelihood of success), if a first attempt at IVF has failed, based on a variety of information that it is fed into a mathematical model. Previous models looked only at the woman's age to make a prediction.

The new model looks at "dozens of factors pertaining to a couple’s fertility, age and health. . . [for example,] drugs given to boost fertility in the mother, the number of embryos with more than four cells reinserted, the woman’s age and weight, the age of the sperm provider and so on. . . . it can apply only to second attempts, since it relies on a history of the first IVF procedure to generate the data needed for the calculation. . . . [The factors include] the number of eggs retrieved [from the first IVF attempt], the number of embryos available for reinsertion [from the first IVF attempt[ and the woman’s previous history of live births."

It is hard to say how much the data will matter to couples. Utimately, the process simply makes what is already a crap shoot before, a somewhat more informed crap shoot where the odds are better known.