There is no evidence that the time from egg retrieval to embryo transfer affects live birth rates in a freeze-all strategy

STUDY QUESTION: Does the time from ovum pick-up (OPU) to frozen embryo transfer (FET) affect reproductive outcomes in a freeze-all strategy? SUMMARY ANSWER: Our study did not detect statistically significant differences between first and subsequent cycles, clinically relevant differences are not ruled out and further and larger studies are required. WHAT IS KNOWN ALREADY: Following controlled ovarian hyperstimulation (COH) delaying FET until the endometrium has returned to an optimal pre-stimulation state may have a significant emotional impact on patients, which adds to the stress and anxiety accompanying a standard IVF cycle. Currently there is no agreement on the best time to perform a FET after a freeze-all cycle in order to maximize reproductive outcomes for the patient. STUDY DESIGN, SIZE, DURATION: Retrospective cohort study of 512 freeze-all cycles, performed between January 2012 and December 2014. COH was performed by either a GnRH antagonist (n = 397) or a long GnRH agonist protocol (n = 115). Ovulation was triggered using either a GnRH agonist (n = 258) or hCG (n = 254). Endometrial preparation was performed in an artificial cycle by either oral (n = 238) or transdermal (n = 274) oestrogen. Differences were considered significant if P < 0.05. PARTICIPANTS/MATERIALS, SETTING, METHODS: Reproductive outcomes between FETs which took place either within the first menstrual cycle following OPU (Cycle 1; n = 263) or afterwards (Cycle = 2; n = 249) were compared. Student's t-test for independent samples, Mann-Whitney U-test and Chi-square analysis were used where appropriate. A multivariable logistic regression analysis was performed adjusting for maternal age, drug used for ovulation trigger, number of retrieved oocytes, number of embryos obtained, day of embryonic development at transfer, number of embryos transferred and type of endometrial preparation. Differences were considered significant if P < 0.05. MAIN RESULTS AND THE ROLE OF CHANCE: Live birth rate (LBR) was significantly higher in FET performed during Cycle 1 vs Cycle >= 2 (37.6% vs 27.3%, respectively; P = 0.01) before adjusting for confounding factors. We found no difference for biochemical pregnancy (49.8% vs 43.8%; P = 0.17), clinical pregnancy (44.1% vs 36.1%; P = 0.07) or pregnancy loss (11.8% vs 16.1%; P = 0.16). A multivariable analysis found no impact of timing of elective FET on LBR (odds ratio, OR 0.73; 95% CI 0.49-1.08). The impact remained not significant after adjusting for number of retrieved oocytes, drug used for ovulation trigger (hCG vs GnRH agonist) and reason for cryopreservation. The factors that significantly affected LBR were: maternal age in both age categories (women between 35 and 40 years vs women below 35 years, OR 0.63, 95% CI 0.4-0.95; and women over 40 years vs women below 35 years, OR 0.34, 95% CI 0.2-0.7), day of embryonic development at transfer (day +4 vs +3; OR 1.7, 95% CI 1.1-2.8) and number of transferred embryos (OR 2.2, 95% CI 1.4-3.3) and oestrogen used for endometrial preparation (transdermal vs oral; OR 0.62, 95% CI 0.4-0.9). LIMITATIONS REASONS FOR CAUTION: The main limitation of our study is its retrospective nature. Although we adjusted our statistical analysis for a number of known and suspected confounders, we cannot exclude the possibility of residual confounding factors. WIDER IMPLICATIONS OF THE FINDINGS: According to our results, clinicians might not need to wait more than one menstrual cycle before performing FET. This allows us to reduce unnecessary delays in FET, without compromising reproductive outcomes.