Prophylactic tranexamic acid for the prevention of postpartum haemorrhage in women with placenta praevia: multicentre, double blind, randomised, placebo controlled, phase 3 trial.

The introduction cites prior research establishing that postpartum haemorrhage is a leading cause of maternal death, that placenta praevia increases risk, and that tranexamic acid reduces haemorrhage in low-risk caesarean deliveries. It explicitly states the gap: 'no evidence exists for tranexamic acid reducing more severe maternal outcomes' and 'previous studies have predominantly enrolled participants from low risk or mixed risk populations'. The rationale logically links this gap to the study's objective. Limitations of prior research (selective enrolment, lack of high-risk data) are acknowledged and addressed by the study design.

Randomization uses a centralised computer-generated sequence with block size six and stratification by site, maternal age, and placenta praevia type. Blinding covers participants, investigators, clinical staff, sponsor, and data analysts. A priori power analysis (80% power, 5% alpha, 20% relative risk reduction) is provided with attrition adjustment. Inclusion and exclusion criteria are clearly listed; the analysis population (ITT with exclusions for consent withdrawal and ineligibility) is defined. Outlier handling is addressed through sensitivity analyses and per-protocol population.

Sex is reported (all female, as the condition is exclusive to pregnant women), and a scientific justification is inherent (placenta praevia only occurs in pregnancy). Age, BMI, and health status (e.g., diabetes, hypertensive disorders) are reported in Table 1. Demographics are extensive: age, BMI, education, mode of conception, parity, number of previous caesarean deliveries, gestational age, and comorbidities. Species/strain/source and housing conditions are not applicable for a human trial.

The trial was approved by the Clinical Research and Applied Ethics Committee of The Third Affiliated Hospital, Guangzhou Medical University on 30 May 2023 (ethics review No 045). Written informed consent was obtained from all patients or their legally authorised representatives. Compliance with the Declaration of Helsinki is not explicitly stated, but the ethics approval process and the statement that the trial was conducted in accordance with ethical principles (implied by the approval) are sufficient. The paper also states that ethical approval from other participating centres was obtained before patient enrolment.

Tranexamic acid (Chengdu Lier Pharmaceutical) and placebo (Hebei Tiancheng Pharmaceutical) are named with location; dose and preparation are described. The statistical software R version 4.4.3 is identified. No antibodies, cell lines, or organisms are used. Therefore, all applicable sub-criteria are adequate.

Primary analysis uses log-binomial mixed effects regression; sensitivity and subgroup analyses are clearly defined. Exact p-value (P=0.01) is given for the primary outcome with 95.2% CI. Secondary outcomes present CIs; exact p-values are not always provided, which is a minor shortfall but does not affect the overall pass. Software is identified. Data presentation includes per-group N, percentages, and CIs, typical for a large trial.

The data availability statement says 'The data underlying the findings in this paper are openly and publicly available and can be found at https://github.com/ChenDunjin-GZHMU'. This GitHub link appears to be a code repository, not a dedicated data repository with a persistent identifier. The statement is vague about the data format and does not provide a DOI or accession number. The code is shared in the supplementary file, which is adequate for code sharing. No repository deposit or accession numbers are provided for the data.

The trial is registered at ClinicalTrials.gov (NCT05811676). Methods are detailed enough for replication, including randomisation, blinding, procedures, and statistical analysis. Limitations are explicitly discussed (generalisability, multiple comparisons, recall bias, single-country setting). Conclusions are proportional: 'statistically significant yet modest reduction' and no overclaiming of causation. Funding sources and competing interests are reported. A reporting guideline (CONSORT) is not explicitly referenced, but the paper follows a standard clinical trial reporting structure.