Related Experiment Videos
Regional one-step and two-step gestational diabetes screening recommendations and birthweight outcomes in Belgium: A
Elena Costa1,2, Christophe de Meester2, An Vercoutere1
1Department of Gynaecology and Obstetrics, Hôpital Universitaire de Bruxelles (HUB), Hôpital Erasme, Université libre de Bruxelles (ULB), Brussels, Belgium.
Background:
Gestational diabetes mellitus (GDM) screening practices remain debated. One-step strategies identify more women with GDM, but their population-level benefits remain uncertain. Belgium provides a natural experiment: Wallonia recommended screening with the one-step strategy based on IADPSG criteria in 2012, whereas Flanders continued to recommend a two-step approach.
Objective:
To compare changes in birthweight-related outcomes and recorded hyperglycaemia in pregnancy (HiP) between these regions after their screening recommendations diverged.
Methods:
We conducted a population-based study including all singleton live births in Wallonia and Flanders during 2010-2011, before the adoption of one-step screening recommendations in Wallonia (n = 211,774), and during 2015-2018 (n = 399,376)., after a transitional implementation period ( between 2012 and 2014). HiP was defined as recorded gestational or pre-existing diabetes. Crude prevalences were calculated by region and period. Logistic and multinomial logistic models estimated changes in HiP, small-for-gestational-age (SGA), large-for-gestational-age (LGA), and macrosomia. Adjusted ratios of odds ratios (aRORs) compared post-versus-pre changes in Wallonia with those in Flanders, accounting for maternal sociodemographic characteristics, body mass index, and parity.
Results:
HiP prevalence increased from 2.74% to 4.29% in Flanders and from 5.95% to 8.52% in Wallonia. However, the relative increase was smaller in Wallonia than in Flanders (aROR 0.92, 95% CI 0.87-0.97). The prevalence of SGA decreased from 6.54% to 5.72% in Flanders and from 8.60% to 8.09% in Wallonia, with a smaller relative decrease in Wallonia (aROR 1.11, 95% CI 1.05-1.15). No differential change between regions was observed for LGA (aROR 1.00, 95% CI 0.97-1.04). Macrosomia remained broadly stable in both regions (9.29-9.16% in Flanders and 6.19-6.51% in Wallonia), although a trivial differential change was observed (aROR 1.05, 95% CI 1.00-1.09).
Conclusions:
Although the absolute increase in HiP prevalence was greater in Wallonia, the relative increase in the odds of HiP was smaller than in Flanders, where two-step screening recommendations were maintained. No consistent improvement in birthweight-related outcomes was observed. These findings suggest that changes in screening recommendations may not translate into clinically meaningful population-level changes, although we cannot exclude challenges in their implementation and uptake in clinical practice.
