Relative infertility: modeling clinical paradoxes.

OBJECTIVES

To assume that a cause of relative infertility will decrease the monthly chance of conception (fecundability) in a dose-dependent manner and, by a mathematical model, to identify common clinical observations and paradoxes that are explainable within this hypothesis.

DESIGN

An empirically based assumption of a population-mean fecundability of 0.2 and the accumulating probability of pregnancy equations and projections were used to examine over time the effects of diminishing such fecundability to one half, one fifth, and one twentieth of normal, and then reversing this effect with ideal treatment at points of 2 years and 5 years.

RESULTS

[1] The duration of infertility is an important and powerful covariate in determining residual fecundability and the chance of pregnancy, with or without treatment. [2] The more substantial the pathology is, the greater should be the likelihood of pregnancy after its effective treatment. [3] Provided no harm is done by treatment, an increase in subsequent fecundability will result whatever the "dose" of the reproductive disturbance, but this will not always mean that pregnancy is probable. [4] The presence of a second infertility factor should compound dramatically the deleterious effects attributable to the first and make it more likely for either factor to be diagnosed.

CONCLUSIONS

Duration of infertility is generally more important than the dose of an infertility factor as a covariate in clinical studies, and more emphasis should be placed on controlling for it. Discouraging clinical reports on the low success of treating certain conditions associated with infertility do not necessarily justify rejecting a hypothesis that such a condition decreases fertility in a dose-dependent manner.