Predicting insulin use among women with gestational diabetes diagnosed in oral glucose tolerance test.

Background: Gestational diabetes mellitus (GDM) is one of the most common complications affecting pregnant women. While most women will achieve adequate glycemic levels with diet and exercise, some will require pharmacological treatment to reach and maintain glucose levels between the desired thresholds. Identifying these patients early in pregnancy could help direct resources and interventions.

Methods: This retrospective cohort of women with GDM diagnosed with an abnormal 75g-OGTT presents data from 869 patients (724 in the diet group and 145 in the insulin group). Univariate logistic regression was used to compare the groups, and multivariable logistic regression was used to identify independent factors associated with the need for insulin. A log-linear function was used to estimate the probability of requiring pharmacological treatment.

Results: Women in the insulin group had higher pre-pregnancy BMI index (29.8 vs 27.8 kg/m, odds ratio [OR] 1.06, 95% confidence interval [CI] 1.03-1.09), more frequent history of previous GDM (19.4% vs. 7.8%, OR 2.84, 95% CI 1.59-5.05), were more likely to have chronic hypertension (31.7% vs. 23.2%, OR 1.54, 95% CI 1.04-2.27), and had higher glucose levels at all three OGTT points. Multivariable logistic regression final model included age, BMI, previous GDM status, and the three OGTT values as predictors of insulin requirement.

Conclusions: We can use regularly collected data from patients (age, BMI, previous GDM status, and the three OGTT values) to calculate the risk of a woman with GDM diagnosed in OGTT needing insulin. Identifying patients with a greater risk of requiring pharmacological treatment could help healthcare services to better allocate resources and offer closer follow-up to high-risk patients.