Clinical Scorecard: What EMERGE Revealed About Early Metformin Beyond Statistical Significance

At a Glance

Key Highlights

• The prespecified primary analysis found a risk difference of −6.9% (95% CI, −15.1% to 1.4%; P = .13) for the composite outcome of insulin initiation or fasting glucose ≥5.1 mmol/L at gestational age 32 or 38 weeks.
• A secondary generalized pairwise comparison analysis reported a win ratio of 1.46 (P = .004); the article characterizes such reanalysis findings as exploratory or suggestive.
• The primary-analysis confidence interval was compatible with effects ranging from substantial benefit to minimal benefit and indicated a low probability of harm on the primary composite outcome.
• With uninformative priors, the estimated posterior probability of any benefit was approximately 95%, and the probability of a reduction exceeding 5 percentage points was approximately 70%.
• The article states that the magnitude and clinical importance of any benefit remain uncertain and that greater certainty requires additional research.

Guideline-Based Recommendations

Diagnosis

Management

• The article reports that early metformin was not superior to placebo for the prespecified primary outcome; it describes the secondary reanalysis as exploratory or suggestive and as a rationale for further research.

Monitoring & Follow-up

Risks

Patient & Prescribing Data

EMERGE randomized 535 pregnant women to placebo or metformin. The article provides no further prescribing or dosing details.

The primary analysis did not reach statistical significance. The article reports evidence more consistent with benefit than harm across the primary analysis and secondary reanalysis, while emphasizing uncertainty about benefit magnitude and clinical importance.

Clinical Best Practices

• The article describes confidence intervals as a way to assess effect magnitude and precision alongside P values, and says clinical relevance should be evaluated rather than relying exclusively on statistical significance.
• The authors state that well-justified secondary reanalyses should be considered exploratory or suggestive and reported fairly and honestly.
• The authors note that investigators may, in some circumstances, prespecify Bayesian analyses alongside frequentist analyses or select win ratios rather than a composite outcome when designing a trial.

Related Resources & Content

• EMERGE randomized clinical trial (2023); citation details not provided in the article text
• CONSORT reporting guideline; citation details not provided in the article text