Reducing the incidence of twins from IVF treatments: predictive modelling from a retrospective cohort.

BACKGROUND IVF treatments carry a high risk of twin pregnancy which confers a higher risk to the mother and child than singletons. Increased use of elective single embryo transfer (eSET) can reduce this twin rate. We aimed to utilize a previously published data set and statistical model based on routinely collected clinical data to predict the outcomes of policies that increase the proportion of eSET. METHODS The models allow simultaneous prediction of outcomes from double embryo transfer (DET) and SET. These models were used to predict outcomes for different scenarios using SET in both the initial (fresh) transfer and over a complete cycle (transfer of all embryos created, with cryopreservation). A total of 16 096 cycles (12 487 fresh and 3609 frozen) from 9040 couples treated between 2000 and 2005 were included in the final analyses. RESULTS For any transfer, SET has about a one-third lower live birth rate relative to DET: this can be partially mitigated by appropriate patient and treatment cycle selection, with several realistic policies performing similarly. However, if we consider complete cycles with embryo cryopreservation, it is possible for repeat SET to produce more live births per egg retrieval than repeat DET. CONCLUSIONS All patients receiving SET would have a higher chance of successful treatment in that cycle if they received DET. The selection of appropriate patients for SET can partially ameliorate the overall loss. For complete cycles, repeat SET could produce more live births per egg retrieval than repeat DET. All treatments involving SET will increase the number of treatments required to achieve a successful outcome and this extra treatment burden will be a significant barrier to the implementation of such treatments.