Abstract
BACKGROUND
Gestational diabetes mellitus (GDM) is associated with adverse pregnancy outcomes, but risk is reduced with identification and early treatment. Glucose measurements are affected by preanalytical sample handling, such as temperature of storage, phlebotomy–analysis interval, and use of a glycolysis inhibitor. We evaluated glucose concentrations and the incidence of GDM after strict implementation of the American Diabetes Association (ADA) preanalytical guidelines, compared with usual hospital conditions.
METHODS
Women screened selectively for GDM at 24–32 weeks' gestation were recruited at their convenience before a 75-g oral glucose tolerance test. Paired samples were taken: the first sample followed ADA recommendations and was transferred to the laboratory on an iced slurry for immediate separation and analysis (research conditions), and the second sample was not placed on ice and was transferred according to hospital practice (usual conditions).
RESULTS
Of samples from 155 women, the mean fasting, 1-h, and 2-h results were 90.0 (12.6) mg/dL [5.0 (0.7) mmol/L], 142.2 (43.2) mg/dL [7.9 (2.4) mmol/L], and 102.6 (32.4) mg/dL [5.7 (1.8) mmol/L], respectively, under research conditions, and 81 (12.6) mg/dL [4.5 (0.7) mmol/L], 133.2 (41.4) mg/dL [7.4 (2.3) mmol/L], and 99 (32.4) mg/dL [5.5 (1.8) mmol/L] under usual conditions (all P < 0.0001). GDM was diagnosed in 38.1% (n = 59) under research conditions and 14.2% (n = 22) under usual conditions (P < 0.0001). The phlebotomy–analysis interval for the fasting, 1-h, and 2-h samples was 20 (9), 17 (10), and 17 (9) min under research conditions and 162 (19), 95 (23), and 32 (19) min under usual conditions (all P < 0.0001). All cases of GDM were diagnosed on fasting or 1-h samples; the 2-h test diagnosed no additional cases.
CONCLUSIONS
Implementation of ADA preanalytical glucose sample handling recommendations resulted in higher mean glucose concentrations and 2.7-fold increased detection of GDM compared with usual hospital practices.
MODELING OF SYNCHRONIZATION OF QUASI-RANDOM AFTERMATH OF A LARGE PERIOD IN CONDITIONS OF EXTERNAL DESTRUCTIVE INFLUENCES
