Background
Frozen embryo transfer (FET) is increasing in prevalence. In contrast to the amount of research performed on the actual cryopreservation procedure, there are limited data with respect to optimal endometrial preparation in FET cycles. Increasingly artificial cycle (AC) preparation is being adopted over the natural cycle (NC) to facilitate greater access to FET. However, there remains a paucity of data comparing pregnancy outcomes between these two commonly used cycle types.

Aims
To examine the efficacy of AC vs NC following FET, by comparing pregnancy outcomes including biochemical, clinical and live birth rates, along with miscarriage rates.

Materials and Method
This is a large single‐centre retrospective analysis, examining a standardised data set from January 2015 to July 2018. It included 3030 cycles (NC = 2033, AC = 997). Main outcomes were biochemical pregnancy (beta‐human chorionic gonadotropin > 5 IU), ultrasound‐diagnosed clinical pregnancy, and live births. Using the χ2 test, the above pregnancy outcomes were compared between AC and NC. A multivariate logistic regression, controlling for factors such as age, embryo quality, and day of blastocyst freeze was further utilised to assess for confounding variables.

Results
No difference was observed between biochemical pregnancy rates (NC = 39.45% vs AC = 37.71%, P = 0.357); statistically significant differences were observed between clinical pregnancy (30.84% vs 26.08%, P = 0.007), and live birth rates (24.40% vs 18.86% P = 0.001). Multivariate analysis confirmed that NC produces superior pregnancy outcomes when controlling for confounding variables.

Conclusion
This analysis demonstrates the non‐inferiority of NC thaw compared to AC, on continuing pregnancy rates. Taken together with patient acceptability and possibly increased obstetric risks with AC, these findings support the use of NC when medically possible.