Impact of new criteria for the diagnosis of gestational diabetes: a maternal and neonatal health outcome and economic analysis in a large tertiary level maternity centre

Abstract

Background: The development of diabetes mellitus during pregnancy, referred to as gestational diabetes (GDM), is associated with adverse maternal and fetal outcomes. Routine screening for this condition is therefore undertaken in all pregnant women. In 2014, gestational diabetes underwent a more liberal diagnostic update, which has been widely adopted in Australia but not universally accepted internationally. Concerns around a rapidly increasing annual incidence and implications for costs and workloads have been raised, but they are assumed to have been offset by improvements in clinical outcomes. As one of the only large Western countries to have adopted universal screening under the new criteria, Australia is uniquely poised to assess the value of such a system change.

Hypothesis: That the change in diagnostic criteria for gestational diabetes has resulted in improved clinical outcomes and equitable health economics.

Aim: To compare an entire cohort of women and babies diagnosed with gestational diabetes in 2014 (under the original criteria) with those diagnosed in 2016 (under the updated criteria) and, in particular, to assess any improvements in outcomes, to attribute costs to the increased incidence and to assess any overall economic benefit.

Methods: All women diagnosed with gestational diabetes in 2014 and in 2016 were included as cases. Control groups in each year were defined as those who underwent screening and had a negative test (pre-existing diabetes was thus an exclusion criteria). Women with multiple pregnancies were excluded from both cases and controls. Demographic data were collected from all groups. Maternal outcomes and fetal outcomes were selected to represent those reported in the studies upon which the new criteria are based. Three broad groups of outcome analyses were undertaken. Firstly, all women in 2014 were compared to all women in 2016 to determine whether the change in policy has caused a hospital-wide improvement in outcomes. Secondly, women with GDM in 2014 were compared to controls in 2014 and women with GDM in 2016 were compared to controls in 2016. Finally, women with GDM in 2014 were compared to women with GDM in 2016. In each analysis, women with GDM were examined as a whole and subdivided into diet-controlled and insulin-controlled. For the economic analysis, models of care for routine pregnancy, GDM diet-controlled and GDM-insulin controlled were costed using average-occasions-of-service for clinical reviews, pharmacy fees for medications and consumables, and Medicare Benefits Schedule item numbers for ultrasound services. Cost-savings were assessed using modelling of adverse outcome avoided based on relative-risk reductions published in the studies upon which the new criteria were based.

Results: There was an increase in annual incidence for GDM from 6.0% to 10.4% with gross costs of care increasing by approximately $900 000 and nett costs of care by approximately $560 000. There was a small hospital-wide reduction in very large babies (>95% for birth weight) from4.31% to 3.61% with no other significant differences between 2014 and 2016. Women with GDM remain a higher risk cohort in both demographics and outcomes than those without GDM, but in 2016 women with GDM that is controlled by dietary measures alone represent a cohort with similar outcomes to women without GDM. Modelling for adverse outcomes avoided by the change in criteria did not reveal a cost-saving in the short-term.

Conclusions: The new criteria for diagnosing GDM has resulted in a marked increase in annual incidence (73% relative, 4.4% absolute) without a significant improvement in maternal and neonatal outcomes. While small numbers of adverse outcomes are likely avoided, it is unlikely the potential short-term savings would outweigh the increase in costs if applying a high-risk model of care to all women with GDM. The new criteria may lead to long-term improvements in health of the mother and/or baby that are cost-effective but further research is required to substantiate this possibility. Future randomized controlled trials into different systems of diagnosis and less expensive models of care for mothers with GDM and their babies are also warranted.