How common is real repeated implantation failure? An indirect estimate of the prevalence

Research question What is the real prevalence of repeated implantation failure (RIF) and what reliable estimates can be given on the risk of false–positive diagnosis after two or three failed IVF attempts. Design A recent theoretical model suggested that commonly used definitions (two or three failed IVF attempts in good-prognosis couples) may expose couples to substantial odds of overdiagnosis and overtreatment. This model, however, was theoretical and based on unproven assumptions that the pregnancy rate in the non-RIF population was 30% and the prevalence of RIF was 10%. In the present study, we applied this model to real data to distinguish the real prevalence of RIF and to provide more reliable estimates on the risk of false–positive diagnosis after two or three failed IVF attempts. To this aim, we retrospectively selected 1221 good-prognosis couples and evaluated pregnancy rates up to the third cycle. Results The clinical pregnancy rate at first, second and third IVF cycle was 52%, 41% and 28%, respectively. A pregnancy rate of 61% was extrapolated in the non-RIF population and 15% among women who had experienced RIF. Therefore, the rate of false–positive diagnoses of RIF after two, three and six failed cycles would be 46%, 25%, and 2%, respectively. Conclusions Our analyses show that estimated prevalence of RIF is 15%. The frequently used definition of RIF based on three failed attempts (but not two) in good-prognosis couples seems justified. Physicians, however, should bear in mind that couples may be inappropriately labelled with this condition in one out of four cases.