Aims: As the incidence of gestational diabetes (GDM) increases there is a move towards nurse specialist led clinics. Our aim was to examine the utility of the HbA1c measurement taken at booking in predicting the pregnancy outcomes of women with GDM, with a view to using the HbA1c measurement to aid triage.
Methods: Observational cohort study of women with GDM by NZ criteria referred to Christchurch Hospital over 18 months. Women without booking bloods and those with multiple pregnancies were excluded. Maternal characteristics and pregnancy outcome data were collected from electronic hospital records. HbA1c measurements were converted to customised centiles (adjusted for gestational week at testing).  Glucose treatment targets were <5.0mmol/l fasting, <7.4mmol/l at one hour.
Results: Included were 446 women with HbA1c measurements taken at a mean (SD) 56 (29.4) days gestation. HbA1c centiles increased with maternal BMI and correlated with glucose values at the diagnostic glucose tolerance test. 94 (21%) of women had a HbA1c ≤10th centile (mean ±SD HbA1c 29 ±1.9mmol/mol) and 166 (37%) had a HbA1c ≤25th centile (mean ±SD HbA1c 30 ±2mmol/mol) at booking. Comparing women with a HbA1c ≤10th or ≤25th centile to those with ≥90th centile (mean ±SD HbA1c 41±2.4mmol/mol): GDM was controlled with diet only in 48.9% or 44.6% vs 11.9% (p<0.001) respectively and emergency caesarean section occurred in 14.9% or 14.5% vs 26.2% (p=0.06) respectively. There was no difference in neonatal outcomes between groups.
Conclusions: HbA1c centiles at booking may be a useful tool when triaging women with GDM with respect to the likely treatment required. Women with lower HbA1c centiles (>1/3rd of our cohort) were more likely to require dietary intervention alone and thus more suitable for initial review in nurse specialist led clinics. We are currently analysing the data to see which of HbA1c centiles vs HbA1c values vs glucose values on the GTT is better for predicting pregnancy outcomes.