ERA After Failed IVF: No Clear Live-Birth Benefit in 1,312 Women

Does an endometrial receptivity assay (ERA) improve the chance of pregnancy or live birth after repeated failed embryo transfers? In a retrospective comparison involving 1,312 women, adjusting transfer timing according to ERA results did not significantly improve either outcome.

The study identified a displaced window of implantation in 45.26% of women who underwent testing. A subgroup with displacement had fewer early miscarriages after personalized transfer, but this observational finding needs prospective confirmation.

When IVF repeatedly fails to produce implantation, patients understandably search for an explanation. If a good-quality embryo has been transferred, could the timing of the endometrium be out of step with the embryo?

ERA was developed around that question. A sample of endometrial tissue is analyzed for gene expression to estimate the window of implantation (WOI), the interval when the lining is thought to be most receptive to an embryo.

Some patients with recurrent implantation failure undergo personalized embryo transfer, moving the procedure earlier or later according to their ERA result.

Whether this precision actually improves pregnancy and birth outcomes is a separate question. A large retrospective study accepted by Frontiers in Endocrinology on September 22, 2026, did not find a significant overall improvement.

A displaced implantation window in about 45% of tested patients

Researchers at the Chongqing Health Center for Women and Children in China evaluated ERA in patients meeting relaxed criteria for recurrent implantation failure between August 2019 and December 2024.

They first analyzed test results from 970 women who had ERA. A displaced implantation window was identified in 45.26%, most commonly a window occurring later than the expected timing.

This suggests that the time when an embryo reaches the uterus and the time when the endometrium is receptive may not align identically in every patient.

That is the rationale behind interest in ERA: if the implantation window varies, identifying it and timing the embryo precisely might improve the chance of pregnancy.

Personalized timing did not significantly improve pregnancy or birth

The investigators then compared outcomes. Of the patients tested, 644 underwent personalized embryo transfer based on ERA results. Their outcomes were compared with 668 patients with recurrent implantation failure who underwent conventional transfer during the same period, giving a total of 1,312 women.

Neither clinical pregnancy nor live birth differed significantly between the ERA-guided and conventional-transfer groups.

The analysis therefore did not establish that finding an individual's implantation window and adjusting transfer timing increased the overall likelihood of pregnancy or having a baby.

This distinction is important. A test's ability to identify a difference in timing is not the same as evidence that using that information improves live birth.

In this study, almost half the tested patients had a displaced window, but detecting it and changing transfer timing did not translate into a significant overall increase in births.

An early-miscarriage signal within the ERA group

There was an interesting finding when the 644 patients who received ERA-guided transfer were examined separately.

Among those with a displaced implantation window whose timing was adjusted, the early miscarriage rate was 7.02%. In those without displacement, it was 19.49%. The difference was statistically significant.

When the analysis was restricted to single-blastocyst transfers, the rates were 5.0% with displacement and 20.76% without displacement.

The investigators suggested that personalized timing might be associated with greater early-pregnancy stability in some patients whose implantation window is displaced.

A crucial limitation remains: patients were not randomly assigned to ERA testing. This was a single-center retrospective analysis of clinical records. The within-ERA subgroup comparison cannot establish that the test itself reduced miscarriage. The authors called for prospective studies to investigate the finding.

Finding a window is different from improving live birth

ERA can classify some patients' endometrial timing as different from the expected pattern. The present study found such displacement in 45.26% of tested women.

But finding a measurable difference does not automatically show that treatment based on that result improves pregnancy or live-birth rates.

This is central to the continuing debate around ERA. Systematic reviews of earlier evidence have likewise not established a consistent improvement in pregnancy outcomes from ERA-guided personalized transfer of euploid embryos.

Repeated implantation failure alone therefore does not establish that every patient needs ERA. At the same time, the observed early-miscarriage difference means a possible benefit for selected subgroups cannot yet be ruled out.

The next question is not simply whether ERA identifies an implantation window, but which patients, if any, gain a meaningful pregnancy or birth benefit from testing.

In this comparison of more than 1,300 women, adjusting transfer timing to ERA results did not provide evidence of an overall improvement in clinical pregnancy or live birth among patients with recurrent implantation failure. More precise information from a test does not necessarily mean better outcomes for patients.

This article is based on the study accepted for publication in Frontiers in Endocrinology on September 22, 2026, titled Clinical Utility of Endometrial Receptivity Assay in Recurrent Implantation Failure Defined by Relaxed Criteria: A Large Retrospective Cohort Study. It is a single-center retrospective study and cannot establish a causal treatment effect of ERA.

Research references

Clinical Utility of Endometrial Receptivity Assay in Recurrent Implantation Failure Defined by Relaxed Criteria: A Large Retrospective Cohort Study. Frontiers in Endocrinology. Accepted September 22, 2026; accepted-article abstract available at verification. DOI: 10.3389/fendo.2026.1991232.