A Prospective Evaluation of Point-of-Care Measurements of Maternal Glucose for the Diagnosis of Gestational Diabetes Mellitus

Abstract

Abstract Background Point-of-care (POC) measurement of glucose is currently recommended only for the monitoring of gestational diabetes mellitus (GDM). This prospective observational study evaluated the use of POC measurements of maternal glucose to diagnose GDM in women being screened selectively with a 1-step 75 g oral glucose tolerance test (OGTT). Methods The strictest preanalytic and analytic international laboratory standards were applied to measure maternal plasma glucose at fasting and at 1 and 2 h post glucose load. The recent International Association of Diabetes and Pregnancy Study Groups diagnostic criteria were used. At the same time, maternal capillary glucose was measured. Because of differences in plasma and capillary glucose measurements, regression analysis of POC capillary glucose results vs laboratory plasma glucose results was conducted. The regression equations for plasma glucose were derived in a derivation cohort (n = 102). These equations were applied in the validation cohort (n = 100). Predicted and actual plasma glucose values were compared. Results Of the 202 women screened, 36.6% were nulliparous, 56.4% were obese, and 81.2% were Irish-born. Two thirds had a single risk factor for GDM, and a third had multiple risk factors. Based on the plasma measurements, 53.5% had GDM. As a predictor of GDM, the diagnostic accuracy of POC measurement was 83.0% (95% confidence interval, 74.2–89.8). Conclusions In high-resource settings where measures to inhibit glycolysis are implemented, the use of POC measurements for the diagnosis of GDM is not justified based on this study. In low- and medium-resource settings, where measures to inhibit glycolysis are not achievable, regression analysis using POC measurements may be acceptable compared with plasma samples subject to glycolysis.