Mode of Evaluation of Results in Artificial Insemination

Abstract

Improved effectiveness in artificial insemination would result from the comparison of studies incorporating different treatment techniques. However, this is difficult to do presently. In studies published to date, the characteristics of the patient populations are different in regard to ethnic background, age distribution, means of selection, etc. To a certain extent, it is possible to minimize or take into account such differences. However, it is essential that the methods used to analyse the data provide indices which have the same meaning. Such a standardized approach has been used in other fields as for example in the evaluation of survival or the effectiveness of contraceptive methods. On the contrary, in artificial insemination diverse indices are currently in use and most are misleading. Even though calculated in the same manner from one study to the other, they do not have the same meaning. This is particularly true for the overall success rate. This index adopted by practically all investigators makes the serious mistake of relating the number of successes to the number of women treated. In fact, this unit, the woman treated, is without meaning when one does not take into account the number of treatment cycles since they can vary greatly for numerous reasons.