Background
Women from rural and remote WA have increased risk for Hyperglycaemia in Pregnancy (HIP), yet poor OGTT completion (50%) and non-adherence to pre-analytical laboratory standards results in significant under-diagnosis (estimated 62%). Measurement of HbA1c and glycated albumin (GA) may improve screening outcomes.
Aims
To determine combined HbA1c and GA cut-points to stratify low- and high-risk for HIP as diagnosed by OGTT (≥24-week gestation) and evaluate accuracy of risk-stratification for detecting adverse birth outcomes.
Methods
Twenty-seven rural and remote WA clinics recruited 694 pregnant women (2015-2018). OGTT were conducted following local pathology guidelines. Paired-sample OGTT time-course comparison of glucose in fluoride/oxalate to fluoride/citrate/EDTA samples informed linear regression correction of OGTT (OGTTc) by delay to analysis (n=12; 363 time-points).
At OGTT, HbA1c (Roche Diagnostics) and GA (AsahiKasei Pharma) were measured and maternal characteristics recorded. Two GP-Obstetricians, blinded to pathology results, defined adverse perinatal outcomes independently as potentially HIP-related.
Outcome measures included: receiver operator characteristics curve derived low-risk (sensitivity ≥95%) and high-risk (specificity ≥90%) cut-points for HbA1c and GA (stratified by BMI: not-obese <30kg/m2; obese ≥30kg/m2), for abnormal OGTTc; OR [95% CI] for composite adverse perinatal outcome (model adjustment: maternal BMI, age, height, ethnicity, and smoking; gestation at OGTT and delivery). To validate thresholds 180 participants were recruited (2020-2022) with OGTT collected into fluoride/citrate/EDTA tubes.
Results
Complete OGTT, HbA1c and GA data was available for 357 participants. Adverse perinatal outcome was common (n =106, 30.7%), however most at-risk women (85.8%) had a normal OGTT (unadjusted OR 1.8 [0.9-3.7], P =0.11 for adverse perinatal outcome). Correction of OGTT by delay to analysis improved identification of risk (unadjusted OR 1.8 [1.1-3.1], P =0.003).
Combined cut-points used to stratify low-risk were HbA1c <4.8% and GA <10.53% (non-obese) or <10.09% (obese); and high-risk were HbA1c ≥5.5% and/or GA ≥12.90% (non-obese) or ≥12.37% (obese). Most women were in the medium risk category (low: 17.6%, medium: 60.5%, high: 21.8%). There was some discordance between abnormal OGTTc and HbA1c-GA risk stratification (low: 14%, medium: 20%, high: 51%). However, the latter was highly predictive of adverse perinatal outcome (47.4% high-risk v 27.6% medium-risk; adjusted OR 2.0 [1.1-3.5] P =0.020) and no abnormal uncorrected-OGTT were missed by low-risk classification. Threshold validation analysis is underway.
Conclusion
As a screening test the OGTT has low sensitivity for identifying women at risk of HIP-related adverse birth outcomes. Combined HbA1c-GA risk-stratification presents an alternative paradigm to detect HIP and reduce the burden of conducting OGTTs.