Design: Cost-effectiveness analysis alongside a randomised controlled trial
Population: 820 women with gestational diabetes at 25 Dutch hospitals
Primary outcome: QALYs, large-for-gestational-age births avoided, and costs
Effect: QALY +0.009 (95% CI 0.004 to 0.014); societal cost −€1004 (−€3503 to €1495)

This economic analysis ran alongside the Dutch SUGAR-DIP trial, in which 820 women with gestational diabetes needing drug treatment were randomised at 16 to 34 weeks to metformin, adding glibenclamide if needed, or to insulin.

Oral agents gave a small QALY gain of 0.009 (95% CI 0.004 to 0.014), about three days in full health. Differences in large-for-gestational-age births avoided (−0.039, −0.097 to 0.018) and societal costs (mean −€1004, −€3503 to €1495) were not statistically significant. The probability that oral agents were cost-effective even at zero willingness to pay was 79%.

The trade-off sits with large-for-gestational-age babies: oral agents were cheaper but slightly less effective on that outcome. For services where insulin, glucose meters and teaching time are costly, as in much of India, the case for starting with metformin is reinforced, provided women understand that some will still need insulin.

Offer metformin as a first-line drug option for gestational diabetes where local guidance allows it.
Explain that a proportion of women on oral agents will still need insulin added.
Discuss the uncertain small difference in large-for-gestational-age births when choosing between the two.
Include glucose meter, strip and teaching costs when comparing options for a woman paying out of pocket.

Why it matters

The choice between metformin and insulin in gestational diabetes is often driven by cost and access, and this gives that choice a trial-based footing.

The statistics, in plain English

The QALY gain is statistically real but clinically trivial. The confidence intervals for costs and large-for-gestational-age births both cross zero, so the trial cannot say that either outcome truly differs. A 79% probability of cost-effectiveness means oral agents usually came out ahead in repeated resampling, not every time.