Analytical approach:
The analysis was based on a decision-analytic model with a lifetime horizon for mothers (56.1 years) and newborns (77.2 years). The authors stated that the analysis took a societal perspective.
Effectiveness data:
Clinical inputs for the model were taken from the literature. It appeared that a selective approach was used and studies might have been known to the authors. Key inputs for the two-hour OGTT based on the new guidelines were taken directly from the Hyperglycemia and Adverse Pregnancy Outcomes (HAPO) study. Hospital discharge databases were used for test accuracy. A meta-analysis was used for the efficacy of gestational diabetes mellitus treatment on pregnancy outcomes. Rates of maternal and neonatal outcomes were key inputs of the analysis.
Monetary benefit and utility valuations:
Utilities were based on published estimates and included both maternal and neonatal endpoints. These were taken from the literature.
Measure of benefit:
Quality-adjusted life-years (QALYs) were used as the summary benefit measure and were discounted at an annual rate of 3%. Maternal and neonatal outcomes were included in the assessment of QALYs.
Cost data:
The economic analysis included costs of treating gestational diabetes (pharmacotherapy, antenatal visits, ancillary diabetes-related visits and ante-partum foetal surveillance), costs of diagnosis (direct costs of materials and analysis as well as indirect costs of travel and opportunity costs for the patient) and costs of selected adverse outcomes (including maternal and neonatal death). Economic data were taken from published sources (details not reported). Cost of treating gestational diabetes mellitus was taken from a USA study published in 2000. Costs were in US dollars ($). The price year was 2012.
Analysis of uncertainty:
One-way, two-way and threshold sensitivity analyses were carried out on probabilities, utilities and costs to assess the robustness of model outcomes. A Monte Carlo simulation was performed using published probability distributions for model inputs. Cost-effectiveness acceptability curves were constructed.