Jovanmandic / iStock
Glucose meters used at the point-of-care (POC) may have limited uses as a diagnostic tool for gestational diabetes mellitus (GDM). A team of researchers found no justification for using POC measurements in diagnosing GDM in high-resource settings. However, this approach might serve a purpose in low- or middle-resource settings without access to accredited laboratories or preanalytical sample handling. Results were published in Clinical Chemistry.
Screening for GDM worldwide is a contentious topic, according to the study’s lead author Eimer O’Malley, BSc, BM, BS, special lecturer and clinical tutor at University College Dublin’s School of Medicine.
In North America, screening is universal. Clinicians use either a two-step 100g oral glucose tolerance test (OGTT) in women >24 weeks gestation or a one-step 2-hour 75g OGTT. Since 2010, a more sensitive 75g OGTT for use in women >24 weeks gestation has been recommended with the emergence of new clinical data. However, this approach is only being used in some countries.
“GDM rates have risen up to 18%, leading to big cost increases for the maternity services through more investigations, more treatments, more interventions,” O’Malley told CLN Stat. Labs also need to follow stricter standards to prevent preanalytical glycolysis of maternal samples.
O’Malley and his colleagues wanted to examine the feasibility of POC testing, which could decrease screening expenses and wait times for results.
To accurately measure plasma glucose, strict preanalytical sample handling is required to avoid glycolysis. “This involves placement of samples collected in a sodium fluoride buffer tube on an ice water slurry and transport to a laboratory within 30 minutes for centrifugation,” O’Malley explained. Such steps may not be feasible in routine clinical practice. Previous studies that compared POC glucose meters with laboratory plasma glucose samples never included measures to inhibit glycolysis, added O’Malley. He and his colleagues sought to compare the diagnostic accuracy of a chosen POC glucose meter with laboratory plasma glucose samples handled under these strict conditions.
More than 200 pregnant women without preexisting diabetes but with risk factors for GDM (based upon the Irish National Guideline for GDM) were selectively screened for GDM at 26-28 weeks’ gestation, using a one step, 2-hour 75g OGTT. The majority were Irish born—more than half were obese and 36.6% were first-time mothers. Two-thirds had a single GDM risk factor, and the rest had multiple risk factors. Investigators collected a venous sample at the time of the OGTT, using measures for sample handling to inhibit glycolysis. “This served as the gold standard reference for the diagnosis of GDM,” said O’Malley. At each time point (fasting, 1-hr and 2-hr), they used the Bayer Contour XT meter to measure the POC capillary sample from a distal fingertip.
The investigators randomly split the cohort into a derivation cohort and validation cohort, using a linear regression method to compare the predicted venous result (based on the POC result) and the actual venous result. Calculations of the sensitivity, specificity, positive and negative predictive values, and accuracy of the POC method took place.
The overall GDM rate in the derivation cohort was 53.5% when strict preanalytical sample handling, along with diagnostic criteria from the International Association of Diabetes and Pregnancy Study Groups was applied. In the validation cohort, the GDM rate was 56% based on laboratory plasma glucose and 51% based on the predicted lab results (derived using the regression equations based on POC values). This corresponded to an accuracy of 83.% for the POC measurements to predict laboratory plasma glucose in this cohort.
Based on these findings, having access to the laboratory in a timely manner and adhering to strict sample handling would be optimal to POC testing to avoid any missed cases, O’Malley said. “If not, such as in a low- or medium-resource setting, then POC testing may be considered as an alternative to laboratory testing.”
Additional work is needed to assess the performance of this method outside of the research setting, such as in other healthcare settings, he said. “Further research is also required using different makes and models of glucose meters and to assess the variability between different meters of the same make and model.”
In a related editorial, David Bruns, MD, Boyd Metzger, MD and David Sacks, MD agreed that POC glucose meters have their limitations in diagnosing GDM. “The diagnostic accuracy with the predicted venous plasma results was not as good as might be hoped,” producing a diagnostic sensitivity of 80.4% and specificity of 86.4%. “Better results might have been obtained with a different meter, but that points to another danger in concluding that glucose meters can be used for diagnosis of GDM: Meters vary in analytical performance,” they noted.
The editorialists focused on a possible alternative to meters: blood tubes containing citrate and fluoride and ethylenediaminetetraacetic acid (CFE tubes) that inhibit glycolysis. An earlier study found that CFE tubes increased diagnostic sensitivity of the OGTT by more than half: 42% to 86%, producing 100% diagnostic specificity. The authors had recommended replacing traditional sodium fluoride-containing tubes with CFE tubes to diagnose GDM.
CFE tubes could work in both high-resource and some resource-limited settings, and offer several financial advantages over glucose meters, the editorialists stated. “The cost of the citrate used in CFE tubes surely is low. Compared with using glucose meters, use of CFE tubes avoids the costs of purchasing glucose meters and reagents, and the expenses associated with training operators and initiating a quality assurance and monitoring program,” they wrote.