Frozen versus fresh embryo transfer in women with low prognosis for in vitro fertilisation treatment: pragmatic, multicentre, randomised controlled trial.
Wei D, Sun Y, Zhao H, Yan J, Zhou H, Gong F, Zhang A, Wang Z, Jin L, Bao H, Zhao S, Xiao Z, Qin Y, Geng L, Cui L, Sheng Y, Sun M, Liu P, Ding L, Liu H, Wu K, Li Y, Lu Y, Xu B, Xu B, Zhang L, Zhang H, Legro RS, Chen ZJ
- DOI
- 10.1136/bmj-2024-081474
- Record issued
- 2026-08-15
- Engine
- 7.39.0
- Exported
- 2026-09-21
Prepared by Alpha1. This document is confidential: it is intended for the recipient it was shared with and must not be redistributed. The live record at alpha1science.com/verify/4e04d23c-c248-4bf1-9eab-95cd3d581ef6 is authoritative.
How this rating was calculated
- IntegrityIntegrity concern ×2−1★
This Kaimen Rigor review uses Kaimen Rigor reviewers trained on a curated corpus of high-fidelity and retracted papers, with expert supervision and curation. It can still make mistakes; verify each finding against the source before relying on it.
This is a well-conducted and transparently reported pragmatic randomised controlled trial. The design is rigorous, with adequate randomization, sample size justification, and pre-specified analyses. Minor reporting gaps include lack of explicit CONSORT adherence and a slight inconsistency in live birth definition.
Both reviewers independently scored all dimensions as pass with high confidence, and their evidence was consistent. The study type is interventional (RCT). Non-applicable sub-criteria (e.g., animal housing, cell line authentication) were excluded. The statistics verification recomputed 3 tests, all consistent; however, this does not confirm the entire statistical analysis.
Numerical inconsistencies
1 finding · worst lowValues that contradict each other or are impossible for the stated sample: recomputed p-values and test statistics, GRIM/GRIMMER checks on summary numbers, percentages against their own counts, totals against their parts, and estimates against their own confidence intervals.
- Internal contradictions in the reported numbersAssessed
Recomputed 3 tests: 3 consistent, 0 inconsistent; 3 via agent-written checks.
- CONSISTENTreported p = .009 · recomputed p = .009Reviewers 1, 2Primary outcome live birth comparison using chi-square test
“132 (32%) of 419 women in the frozen embryo transfer group had a live birth compared with 168 (40%) of 419 in the fresh embryo transfer group (P=0.009)”
Taken as given: The numbers 132 and 168 are the event counts in each group.; The numbers 419 and 419 are the total participants in each group.; The test used is Pearson's chi-square test without continuity correction.Method: Two-tailed Pearson chi-square test on the 2x2 table.How we recomputed it: pChi2x2(132, 419-132, 168, 419-168) - CONSISTENTreported p = .020 · recomputed p = .021Reviewers 1, 2Clinical pregnancy comparison using chi-square test
“The rate of clinical pregnancy was also lower in the frozen versus fresh embryo transfer group (39% (164 of 419) v 47% (197 of 419); difference between groups −7.9% (−14.6% to −1.2%); relative ratio 0.83 (0.71 to 0.97); P=0.02)”
Taken as given: The numbers 164 and 197 are the event counts in each group.; The numbers 419 and 419 are the total participants in each group.; The test used is Pearson's chi-square test without continuity correction.Method: Two-tailed Pearson chi-square test on the 2x2 table.How we recomputed it: pChi2x2(164, 419-164, 197, 419-197) - CONSISTENTreported p = .040 · recomputed p = .038Reviewer 1Cumulative live birth comparison using chi-square test
“The cumulative live birth rate of embryo transfers within one year of randomisation was still lower in the frozen embryo transfer group than that in the fresh embryo transfer group (44% (185 of 419) v 51% (215 of 419); difference between groups −7.2% (−13.9% to −0.4%); relative ratio 0.86 (0.75 to 0.99); P=0.04)”
Taken as given: The numbers 185 and 215 are the event counts in each group.; The numbers 419 and 419 are the total participants in each group.; The test used is Pearson's chi-square test without continuity correction.Method: Two-tailed Pearson chi-square test on the 2x2 table.How we recomputed it: pChi2x2(185, 419-185, 215, 419-215)
- lowinternal contradictionThe abstract reports live birth rate as 32% (132 of 419) and 40% (168 of 419), but Table 4 reports 31.5% and 40.1% for the same numbers. The percentages are slightly inconsistent due to rounding, but the counts are identical.
Abstract: '32% (132 of 419) v 40% (168 of 419)' ; Table 4: '132 (31.5) | 168 (40.1)'
Table 4reviewer’s wording - lowinternal contradictionThe definition of live birth in the methods states '≥28 weeks', but a footnote in Table 4 mentions two patients with live birth at 27 weeks were included. This is a minor inconsistency in the application of the definition.
Methods: 'Live birth was defined as the delivery of a neonate with heart beat and breath at ≥28 weeks of gestation.' ; Table 4 footnote: 'Two patients who had a live birth at 27 weeks of gestation were included in the numerator for the intention-to-treat analysis of live births.'
Table 4reviewer’s wording
Overstated conclusions
None foundConclusions that reach past what the paper's own results support — including a significance claim that no longer holds when the statistic is recomputed, and efficacy resting on an unvalidated surrogate endpoint.
Checked — nothing surfaced.
4 major claims checked against the paper's own evidence: all adequately supported.
- supportedReviewers 1, 2Frozen embryo transfer results in a lower rate of live birth compared with fresh embryo transfer in women with low prognosis.The primary outcome analysis directly supports this claim with a statistically significant difference.Evidence: Primary outcome: 132/419 vs 168/419, RR 0.79 (95% CI 0.65-0.94), P=0.009.
“the rate of live birth was lower in the frozen embryo transfer group than in the fresh embryo transfer group (32% (132 of 419) v 40% (168 of 419); relative ratio 0.79 (95% confidence interval 0.65 to 0.94); P=0.009).”
AbstractFind in source - supportedReviewers 1, 2The cumulative live birth rate is lower in the frozen embryo transfer group.The secondary outcome analysis supports this claim with a statistically significant difference.Evidence: Cumulative live birth: 185/419 vs 215/419, RR 0.86 (95% CI 0.75-0.99), P=0.04.
“The cumulative live birth rate was lower in the frozen embryo transfer group compared with the fresh embryo transfer group (44% (185 of 419) v 51% (215 of 419), 0.86 (0.75 to 0.99)).”
AbstractFind in source - supportedReviewers 1, 2Fresh embryo transfer may be a better choice for women with low prognosis in terms of live birth rate.The conclusion is proportional to the primary outcome result, though the word 'may' appropriately reflects the need for further research.Evidence: Primary outcome result and discussion.
“Fresh embryo transfer may be a better choice for women with low prognosis in terms of live birth rate compared with a freeze-all strategy.”
ConclusionFind in source - supportedReviewer 1The findings do not support the routine use of freeze-all strategy in women with low prognosis.The primary outcome showing lower live birth with frozen transfer supports this claim.Evidence: Primary outcome result.
“The findings do not support the routine use of freeze-all strategy in women with low prognosis for IVF”
What this study addsFind in source
Efficacy claim is anchored to an adequate endpoint and a meaningful effect.
- ADEQUATESurrogate endpointThe primary outcome is live birth, a hard clinical outcome, not a surrogate.
“The primary outcome was live birth, defined as the delivery of neonates with a heartbeat and respiration at ≥28 weeks’ gestation.”
- ADEQUATEEffect sizeThe effect size is a 8.6% absolute reduction in live birth rate (32% vs 40%), which is clinically meaningful and statistically significant.
“the rate of live birth was lower in the frozen embryo transfer group than in the fresh embryo transfer group (32% (132 of 419) v 40% (168 of 419); relative ratio 0.79 (95% confidence interval 0.65 to 0.94); P=0.009)”
Data authenticity concerns
None foundAn adversarial read for patterns associated with data that may not be genuine: results that look too clean, implausibly large effects, duplicated data or images, and methods that do not match the results reported.
Checked — nothing surfaced.
Reporting gaps
None foundRequired detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.
Checked — nothing surfaced.
The introduction cites prior research on freeze-all strategies in normal/good prognosis women and notes the lack of evidence in low prognosis women. The hypothesis follows logically from the cited evidence and the authors' previous trials. Limitations of prior research (observational studies with inconsistent results) are acknowledged.
“Randomised trials found that compared with fresh embryo transfer, the freeze-all strategy yielded a comparable or higher rate of live birth in women with normal or good prognosis. Whether women with low prognosis benefit from the freeze-all strategy is, however, unclear.”
“Based on our previous trials and the possible adverse effect of superovulation on the endometrium, we tested the hypothesis that frozen embryo transfer would result in a higher rate of live birth compared with fresh embryo transfer in women with a low prognosis.”
“Current evidence on the efficiency of the freeze-all strategy compared with fresh embryo transfer in women with a low prognosis is primarily from observational studies with inconsistent results.”
“Current evidence on the efficiency of the freeze-all strategy compared with fresh embryo transfer in women with a low prognosis is primarily from observational studies with inconsistent results.”
“Based on our previous trials and the possible adverse effect of superovulation on the endometrium, we tested the hypothesis that frozen embryo transfer would result in a higher rate of live birth compared with fresh embryo transfer in women with a low prognosis.”
Randomization was centralized and stratified by age and site, with the sequence inaccessible to recruiters. Blinding was not feasible in this pragmatic trial, but the open-label design is inherent to the intervention and acknowledged. A priori sample size calculation is provided. Inclusion/exclusion criteria are clearly defined. The analysis populations (ITT, per-protocol, per-treatment) are pre-specified, addressing missing data and crossovers.
“The randomisation sequence was generated by the data-coordinating centre in Shandong University and input into the online central randomisation platform ( http://www.medresman.org ), which was inaccessible to the researchers responsible for recruitment and enrolment.”
“With 80% power at a two-sided significance level of 0.05, we determined that the minimum total sample size was 713. Taking into account a 15% drop-out rate, we planned to enrol 838 women.”
“In the secondary analyses, as prespecified in the study protocol, we performed per protocol analysis in women who complied with the study protocol, excluding those with protocol deviations, and per treatment analysis according to the actual treatment that women received.”
“The randomisation sequence was generated by the data-coordinating centre in Shandong University and input into the online central randomisation platform ( http://www.medresman.org ), which was inaccessible to the researchers responsible for recruitment and enrolment.”
“With 80% power at a two-sided significance level of 0.05, we determined that the minimum total sample size was 713.”
“After randomisation, both women and their doctors were informed about the assignment.”
The trial reports age, BMI, duration of infertility, and ovarian reserve markers (AFC, AMH, FSH, LH, estradiol) in Table 1. Both sexes are not applicable as participants are women. Demographics are limited to age and BMI, but these are the relevant biological variables for this fertility trial.
“AMH (pmol/L) | 10.9 (7.2-16.9) | 11.6 (7.3-17.8)”
“Age (years) | 34.0 (31.0-37.0) | 33.0 (31.0-37.0)”
“Duration of infertility (years) | 3.0 (2.0-5.0) | 3.0 (1.5-5.0)”
The paper states approval by the ethics committee of Hospital for Reproductive Medicine Affiliated to Shandong University (ID: 2020-38) and all participating centres. Written informed consent was obtained from all patients. Regulatory compliance is implied by adherence to standard trial conduct, though not explicitly named.
“The trial was approved by the ethics committee of Hospital for Reproductive Medicine Affiliated to Shandong University (ID: 2020-38) and all participating centres.”
“All patients provided written informed consent.”
“The trial was approved by the ethics committee of Hospital for Reproductive Medicine Affiliated to Shandong University (ID: 2020-38) and all participating centres.”
“All patients provided written informed consent.”
The trial uses named medications with manufacturers (Crinone; Merck Serono, Duphaston; Abbott, Progynova; Delpharm Lille). Software used for analysis (SAS 9.4) and sample size (PASS 14.0) are identified. No antibodies, cell lines, or organisms are applicable.
“luteal phase support was initiated with vaginal progesterone gel (Crinone; Merck Serono) 90 mg daily and oral dydrogesterone (Duphaston; Abbott) 10 mg twice daily.”
“SAS version 9.4 was used for all analyses.”
“luteal phase support was initiated with vaginal progesterone gel (Crinone; Merck Serono) 90 mg daily and oral dydrogesterone (Duphaston; Abbott) 10 mg twice daily.”
“SAS version 9.4 was used for all analyses.”
Tests are named (χ², Fisher's, t-test, Wilcoxon rank sum). Normality was tested. Exact p-values are reported for primary and secondary outcomes. Effect sizes (relative ratios) with 95% CIs are provided. Software (SAS 9.4) is identified. Data presentation includes per-group n and percentages. Mathematical plausibility checks on reported percentages and counts are consistent.
“the between group differences tested by χ 2 test or Fisher’s test with <5 expected frequency.”
“the between group differences tested by χ 2 test or Fisher’s test with <5 expected frequency.”
“132 (32%) of 419 women in the frozen embryo transfer group had a live birth compared with 168 (40%) of 419 in the fresh embryo transfer group (P=0.009)”
“relative ratio of 0.79 (95% CI 0.65 to 0.94)”
The data availability statement provides a direct link to a GitHub repository containing the data and code. This is a concrete access route, satisfying the requirement. The repository is public, so repository_deposit and accession_numbers are not applicable in the traditional sense but the GitHub link serves as the repository.
“The data underlying the findings in this paper are openly and publicly available and can be found here: https://github.com/hepingzhangyale/FreFroLowProg.git”
“The data underlying the findings in this paper are openly and publicly available and can be found here: https://github.com/hepingzhangyale/FreFroLowProg.git”
Trial registration number is provided (ChiCTR2100050168). Methods are comprehensive. Limitations are explicitly discussed, including the pragmatic design and embryo transfer differences. Conclusions are proportional to the evidence. Funding sources and competing interests are declared. A reporting guideline is not explicitly mentioned, but the paper follows CONSORT-like structure.
“Trial registration Chinese Clinical Trial Registry ChiCTR2100050168.”
“Our study has some limitations. Firstly, as a pragmatic trial, we did not standardise the stimulation protocol, the number or stage of embryos for transfer, or the regimen for endometrial preparation in the frozen embryo group but followed clinical routines in study sites.”
“Trial registration Chinese Clinical Trial Registry ChiCTR2100050168.”
Registered (1 ID: Chinese Clinical Trial Registry). No reporting guideline cited.
Broken references and links
None foundReferences checked against Crossref, OpenAlex and Retraction Watch for retractions and resolvability, plus declared data and code links probed for whether they resolve to content matching the paper.
Checked — nothing surfaced.
Checked 41 references by DOI: 41 verified.
Every extracted reference resolved against Crossref/OpenAlex with no retraction flags.
1 data/code link checked; 1 live.
- codeGitHubLIVEHTTP 200https://github.com/hepingzhangyale/FreFroLowProg.gitResolves to GitHub (code repository).
Copyediting
5 minorWording, consistency and formatting errors that need correcting before submission.
No major wording or formatting errors. 5 minor suggestions below.
5 copyedit issues flagged: mostly typo, consistency, clarity.
- MINORtypoAbstract, Results“relative ratio 0.79 (95% confidence interval 0.65 to 0.94); P=0.009”→ Consider using 'relative risk' or 'risk ratio' consistently instead of 'relative ratio'.Terminology inconsistency.
- MINORconsistencyTable 4, footnote“Two patients who had a live birth at 27 weeks of gestation were included in the numerator for the intention-to-treat analysis of live births.”→ Clarify the definition of live birth in the methods to reconcile with this footnote.Potential inconsistency in live birth definition.
- MINORclarityMethods, Statistical analysis“For the crossovers, the pregnancy outcomes of the first embryo transfer of the indexed cycle were included in the intention-to-treat analysis according to their originally randomised groups.”→ Consider rephrasing for clarity.Sentence is slightly convoluted.
- MINORtypoAbstract, Results“relative ratio 0.79”→ Consider using 'relative risk' or 'risk ratio' for consistency with standard terminology.The term 'relative ratio' is unusual; 'risk ratio' is more standard.
- MINORconsistencyTable 4, footnote“Two patients who had a live birth at 27 weeks of gestation were included in the numerator for the intention-to-treat analysis of live births.”→ Clarify the discrepancy with the definition of live birth at ≥28 weeks.This footnote contradicts the primary outcome definition; should be explained.
The published work is robust and well-reported. An informed reader should weigh the minor inconsistencies (live birth definition, rounding of percentages) and the lack of explicit CONSORT adherence, but these do not undermine the main conclusions. No erratum is warranted for the identified issues, though clarifying the live birth definition would improve transparency.
- 1.HIGHreportingIn the Methods or a dedicated section, explicitly state adherence to CONSORT reporting guidelines and provide the checklist as supplementary material.Both reviewers noted that reporting guideline adherence is not explicitly stated, which is a common expectation for RCTs.
- 2.HIGHreportingClarify the definition of live birth in the Methods to reconcile with the Table 4 footnote that includes two patients with live birth at 27 weeks.The copyedit and integrity checks flagged an internal contradiction between the stated ≥28 weeks definition and the inclusion of 27-week births.
- 3.MEDIUMcopyeditReplace the term 'relative ratio' with 'risk ratio' or 'relative risk' consistently throughout the manuscript.The copyedit flagged 'relative ratio' as non-standard terminology; consistency improves clarity.
- 4.MEDIUMreportingIn the Ethics statements, explicitly state compliance with the Declaration of Helsinki or other regulatory framework.Reviewer 1 noted that regulatory compliance is implied but not explicitly named; adding this strengthens the ethics reporting.
- 5.MEDIUMdata codeAdd a data dictionary or codebook to the GitHub repository to enhance usability of the public dataset.Reviewer 2 suggested this to improve reproducibility and data usability.
- 6.MEDIUMreportingAdd a statement about the availability of the full study protocol, if not already in supplementary materials.Reviewer 2 suggested this to enhance transparency.
- 7.LOWcopyeditRephrase the sentence in Methods, Statistical analysis about crossovers for clarity.The copyedit flagged the sentence as slightly convoluted.
- 8.LOWreportingIn the Discussion, explicitly address the lack of blinding as a limitation and its potential impact on outcomes.Reviewer 1 suggested this to strengthen the limitations section.
- 9.LOWstatisticsClarify how missing data for laboratory tests (e.g., FSH, LH) were handled in the analysis.Reviewer 1 suggested this to enhance statistical transparency.
- 10.LOWreportingIn the Results, consider reporting the number of participants with missing data for each outcome.Reviewer 1 suggested this to enhance transparency.
- 11.LOWdata codeProvide the full statistical analysis code with comments in the supplementary material or repository.Reviewer 1 suggested this to facilitate reproducibility.
- 12.LOWreportingIn the Discussion, expand on the generalizability of the findings to other populations and settings.Reviewer 1 suggested this to strengthen the discussion.
The star rating is the report’s one-glance summary. Every paper starts at 5★ and loses stars for the concrete problems the review finds — so a rating is never a vague average, it’s a running total you can read line by line under “How this rating was calculated.”
- Reporting — 8 dimensionseach dimension that fully fails−½★
- each dimension partially met−¼★
- Statistics · Integrity · Claimseach serious problem−1★
- each medium problem−½★
- Citationseach retracted or unverifiable reference−¼★
- Copyeditonly when the manuscript needs a full edit−½★
The rating never drops below 1★, and a demonstrable critical failure (an impossible statistic, a proven ethics violation) caps it at 1★ on its own — so the stars can never look healthy when the verdict is CRITICAL.
The rating draws on a panel of agents. Three independent Kaimen Rigor reviewers grade the eight dimensions below across several independent passes (the shown verdict is their majority vote — steadier than any single run), isolate the paper’s major claims and check its own evidence backs them, and flag integrity concerns. Alongside them, a citation agent resolves every reference against Crossref, OpenAlex, and Retraction Watch; a statistics agent recomputes reported tests; and rule-based checks verify that declared data/code links actually resolve. Full text is required — an abstract-only submission is not analyzed.
Graded against NIH, MDAR, ARRIVE 2.0, CONSORT, EQUATOR, and RRID guidelines. A dimension that doesn’t apply to the study type is skipped, never penalized.
This Kaimen Rigor review is model-assisted and is not a substitute for formal expert review. It complements human evaluation by surfacing potential methodological concerns — verify each finding against the source.