Modelling the impact of single embryo transfer in a national health service IVF programme

Greater use of single embryo transfer (SET) to reduce twin rates associated with IVF requires good information on prognostic factors and appropriate models of treatment outcomes. Using data from a cohort of 1198 IVF cycles, we have developed a statistical model of live birth and twin outcomes in terms of routinely measured clinical parameters. From this model, we predict potential outcomes if those who had two embryos transferred had actually received SET. Embryo quality, age, FSH level, idiopathic diagnosis, sperm count, smoking and alcohol consumption are all significant factors predicting outcome. Couples with good embryos and good prognosis have a much greater risk of producing twins. In this cohort, to achieve a 10% twin rate would require 55% SET which, without selection of appropriate cycles, would lead to a reduction in success rate from ca. 21% to 17%. Selecting on the basis of twin risk can partially mitigate this reduction to give a success rate of 18.5%. The use of SET to reduce twin rates will lead to a significant reduction in treatment success. Around half this reduction could be mitigated with careful selection of patients and cycles, including embryo quality.