Is Stratification of "Low-Risk" Women with Gestational Diabetes Mellitus to Usual Antenatal Care Safe?

Abstract

Introduction: Gestational Diabetes Mellitus (GDM) patients are stratified into low-risk and high-risk groups in Canberra, Australia, according to whether their glycaemic control reaches the target levels with lifestyle measures only. High-risk patients, in whom glycaemic control is unsatisfactory, are referred to a multidisciplinary “diabetes in pregnancy” team, while low-risk patients continue regular antenatal care. The aims of this study were to test the accuracy of the current stratification system of GDM treatment in Canberra, and to access whether low-risk patients have satisfactory perinatal outcomes compared to the high-risk patients, considering their less intensive antenatal care. Methods: A retrospective clinical audit of GDM patients treated between 01/01/2010 and 30/06/2014 was conducted. Maternal demographic data and neonatal/maternal clinical outcomes data were analysed including, for key outcomes, comparison with outcomes for the background population in the ACT. Results: Low-risk (n=509) compared to high-risk (n=466) GDM mothers were younger (31.7±4.8 vs 32.6±5.3 years-old, p=0.009), leaner [body mass index (BMI) 26.3±6.7 vs 29.3±7.5 kg/m2, p<0.001], and less parous (0.73±1.0 vs 0.98±1.2 times, p<0.001), with less past GDM (13.2% vs 23.2%, p<0.001), less family history of diabetes (55.4% vs 67.0%, p=0.001), and a lower fasting glucose level in the oral glucose tolerance test (OGTT) (4.9±0.5 mmol/l vs 5.0±0.8 mmol/l, p<0.001). There were more South-East Asian women in the low-risk group (19.4% vs 11.9%, p=0.002). Low-risk mothers had lower rates of pregnancy-induced hypertension (PIH) (6.1% vs 11.8%, p=0.002; ACT 5.7%), induced labour (23.2% vs 50.6%, p<0.001) and elective Caesarean-section (CS) (14.1% vs 20.4%, p=0.010). Rates of emergency CS were similar in the low- and high-risk groups (16.7% vs 19.1%, p=0.328; ACT 14.9%). The rate of preterm delivery (delivery before 37 weeks gestation) was higher in the low-risk group, (9.8% vs 6.0%, p=0.014; ACT 8.3%), attributed to a higher rate of spontaneous preterm delivery (6.1% vs 2.6%, p=0.010). After adjusting for maternal age, BMI, parity, smoking status and alcohol consumption during pregnancy, premature delivery was still more likely in the low-risk group (odds ratio 1.897, 95% Confidence Interval 1.137-3.164). For neonatal outcomes, there were no differences in rates of babies with birth weight >4000g (5.5% vs 7.1%, p=0.309; ACT 11.8%), shoulder dystocia (1.6% vs 1.5%, p=0.930), hypoglycaemia (6.1% vs 7.1%, p=0.532), respiratory disorder (6.3% vs 6.0%, p=0.857), and hyperbilirubinaemia (8.8% vs 10.7%, p=0.320). There was a trend towards a lower rate of customized large for gestational age infants (cLGA) in the low-risk group, compared to the high-risk group (6.1% vs. 9.4%, p= 0.050). The rate of neonatal admission to the intensive care unit (NICU)/special care nursery (SCN) was higher in the low-risk group (16.7% vs 10.9%, p=0.010; ACT 14.7%). However, this difference might have been attributed to the different NICU/SCN admission criteria adopted by the two evaluated hospitals. Conclusion: The stratification system is efficient: low-risk compared to high risk patients were younger, leaner, and had less past GDM, less family history of diabetes and lower fasting glucose during the OGTT. Adverse pregnancy outcomes were either less (PIH, delivery interventions, cLGA) or similar (emergency CS and some neonatal complications) in the low compared to high risk group. One exception was a higher rate of preterm delivery among low-risk women. Some adverse neonatal outcomes for low-risk women were also higher than in the general ACT population. The treatment pathway of low-risk GDM patients has considerable merit, but requires further assessment and optimisation to ensure safety.