Thinking about getting pregnant through in vitro fertilization (IVF)? Then the time couldn’t be better, as Stanford University researchers have developed a way of predicting a woman’s chances of becoming pregnant after IVF treatment.

In IVF treatment, a number of eggs harvested from the woman are fertilized in the lab with the man's sperm. The fertilized eggs are allowed to grow for about 5 days. Then, the ones that are deemed most viable are selected for implanting into the woman's uterus.

IVF is both an expensive and emotional procedure with success only in 18 to 45 percent of the cases, according to the Society for Assisted Reproductive Technology.

The new method can predict with 70-percent accuracy whether a current IVF procedure will result in pregnancy.

Dr. Mylene Yao, professor of obstetrics and gynecology at Stanford and lead author of the study along with her colleagues analyzed data from 665 IVF cycles performed at the university in 2005 to determine link between IFV outcomes and 30 variables in patient characteristics, clinical diagnoses, treatment protocol and embryo characteristics.

The researchers found that four of the 30 factors were critical in predicting IVF pregnancy. The factors were as follows:1. the total number of embryos produced during the procedure;2. the number of embryos that have reached the eight-cell stage development;3. percentage of embryos that stopped dividing and would die;4. the woman’s follicle-stimulating hormone level – an estimate of ovarian function.

The study found that these four factors were better at predicting pregnancy when combined.

“Using only these four factors, we can predict pregnancy with an accuracy of 70%,” Dr. Yao said in the study, adding that knowing them may prove “critical in counseling patients,” who have already attempted IVF and are wondering if they should try again. Also, the method might “improve treatment” and ultimately will help with “developing…more customized treatments.”

The findings of the study were published in the July 2 issue of PLoS ONE.