Evaluation of point-of-care maternal glucose measurements for the diagnosis of gestational diabetes mellitus.

Abstract

Using updated laboratory standards as the reference, we aimed to compare point-of-care (POC) maternal capillary glucose testing with the diagnostic accuracy of reference and customary venous samples. Women screened selectively with a one-step 75-g oral glucose tolerance test (OGTT) at 24-28 weeks' gestation were conveniently recruited to this prospective observational study. Two venous samples and one capillary sample were taken at each OGTT time point. Venous sample one was a fluoride-EDTA (FE) tube placed on an ice-slurry until cell separation and analysis within 30 minutes (reference standard). Venous sample two was transported in a tube containing FE (without ice) (customary practice). A capillary sample was used for POC testing. Various cut-off points for the POC sample were examined to evaluate diagnostic accuracy. The sensitivity, specificity, positive and negative predictive values and accuracy of POC capillary glucose for the diagnosis of GDM. Of 108 women, GDM was detected in 47.2% (n = 51), 17.6% (n = 19) and 24.1% (n = 26) using the reference standard, customary practices and POC, respectively (P < 0.001). However, based on adjustment of the POC fasting diagnostic threshold from ≥5.1 to ≥4.8 mol/l (aPOC), sensitivity, specificity, PPV, NPV and accuracy improved to 92.5, 76.5, 69.8, 94.5 and 94.5%, respectively. POC capillary maternal glucose tests were superior to customary laboratory practices for diagnosing GDM. This has considerable potential, particularly in healthcare settings where facilities for phlebotomy are distant from the laboratory or pre-analytical sample handling is substandard. Adjusted point-of-care glucose measurements have potential in the diagnosis of gestational diabetes mellitus.