Heterogeneity, the masked part of reproductive technology success rates

Heterogeneity in women and men's fecundity is a well-established fact. The selection process of men and women treated for infertility might bias the success rates of reproductive technology. Bias may also arise from frequency and timing of intercourse with respect to the day of ovulation.

Several datasets were collected and analysed. They concern normally fertile couples who used natural family planning methods and infertile couples treated by artificial insemination with husband or donor's spermatozoa. The effects of heterogeneity on the success rates of treatment cycles are described and solutions are proposed as to data collection and statistical analysis in this specific field.

The decrease in the success rates along successive cycles of assisted reproduction is a consequence of heterogeneity. The probability of conception varies among women and among men. After a first success, the probability of another success is higher. After a failure, the probability of success is lower. The most specialised centres treat the less fecund couples. There is a negative correlation between men and women's fertility in case of oligo/azoospermia. The fecund window cannot be correctly located by calendar calculations, but more appropriately by assessment of cervical mucus at the vulva. The variability of this location is wide. The decrease in the success rate with men and women's age results from a complex mixture of an increase in the proportion of sterile patients and a decrease in fecund patients' fecundity.

Care should be taken to limit the bias due to patient selection and specific statistical methods should be used to allow for the progressive selection of patients during fertility studies and for the variability of the frequency and the timing of intercourse or insemination relative to ovulation.