Paper Spotlight is a new EUSOM section in which we pick a recent paper that matters for obstetric medicine and discuss what it means for daily practice.
Our first pick is the economic evaluation of the Dutch SUGAR-DIP trial, just published open access in BJOG: oral glucose-lowering agents (metformin, with glibenclamide added if needed) versus insulin for gestational diabetes, from a societal and healthcare perspective.
Paper: Broulikova HM, van der Wel AWT, Rademaker D, et al., Bosmans JE, Painter RC, for the SUGAR-DIP Study Group. Oral antihyperglycemic agents versus insulin for gestational diabetes mellitus: societal cost-effectiveness analysis of the SUGAR-DIP randomized controlled trial. BJOG 2026. doi:10.1111/1471-0528.70347
Why this question matters
Gestational diabetes affects about one in seven pregnancies worldwide, and many women need medication once diet alone falls short. Insulin has long been the default, but metformin is increasingly used first-line: it is oral, cheaper, and preferred by many patients. Guidelines still differ, which makes trial-based evidence on quality of life and costs valuable.
The original SUGAR-DIP trial (JAMA, 2025) did not show non-inferiority of the oral strategy for large-for-gestational-age (LGA) infants. This follow-up asks whether the two strategies differ in what women experience and what care costs.
What they did
- Design: cost-effectiveness analysis alongside a multicentre, open-label RCT in 25 Dutch hospitals.
- Population: 820 women with a singleton pregnancy and diet-insufficient GDM at 16–34 weeks, randomised to the oral strategy (n=409) or insulin (n=411).
- Oral strategy: metformin first, glibenclamide added if needed, and insulin replacing glibenclamide if still needed.
- Outcomes: LGA prevented and quality-adjusted life-years (QALYs, from EQ-5D-5L at inclusion, 36 weeks and 4–6 weeks postpartum).
- Costs: healthcare, informal care and lost productivity, up to 6 weeks postpartum, in 2022 euros.
What they found
The oral strategy gave slightly better quality of life at a similar or lower cost, with a non-significant increase in LGA.
| Outcome (oral vs insulin) | Difference (95% CI) |
|---|---|
| QALYs gained | +0.009 (0.004 to 0.014), roughly 3 days in full health |
| LGA prevented | −0.039 (−0.097 to 0.018), i.e. about 4% more LGA with oral agents |
| Total societal costs per woman | −€1,004 (−€3,503 to €1,495) |
| Healthcare costs only | +€54 (−€1,215 to €1,324) |
| Probability cost-effective at €0 WTP | 79% for both outcomes |
Most of the societal saving came from less lost productivity in the oral group. In the post hoc subgroup controlled on metformin alone (55% of the oral arm), LGA was 19.7% versus 19.9% with insulin, QALYs were again higher (+0.009), and societal costs were about €1,070 lower.
Our take
This is the first trial-based economic evaluation of drug treatment for GDM, and it strengthens the case for metformin as the first pharmacological step. More than half of women reached good glucose control on metformin alone, with LGA rates equal to insulin, better quality of life and lower costs.
The weak spot of the sequence seems to be the next step. Earlier SUGAR-DIP results linked glibenclamide to more maternal and neonatal hypoglycaemia, and the excess LGA in the oral arm was not seen in the metformin-only group. In practice: start with metformin, and when it is not enough, add or switch to insulin rather than reach for glibenclamide.
Three caveats keep us honest:
- The metformin-only finding is a post hoc subgroup, defined after randomisation and sensitive to covariate adjustment.
- The QALY gain is real but small, below usual thresholds for a clinically important difference.
- Follow-up stopped at 6 weeks postpartum. Long-term effects of LGA and in utero metformin exposure on child metabolic health are not captured.
Take-home for the clinic
- Metformin is a sound first-line drug for GDM: effective for most women, well accepted and cost-saving from a societal view.
- When metformin is not enough, insulin is the logical next step.
- Discuss with women the trade-off between oral convenience and a possible small rise in LGA.
Question for our members: what is your next step after metformin failure — glibenclamide, insulin, or something else? Share your practice with us via LinkedIn or at the next EUSOM webinar.
Written by Koen Verdonk for EUSOM. The author was a site investigator in the SUGAR-DIP Study Group.