Evaluation of fasting and random plasma glucose for diagnosis of gestational diabetes

ABSTRACT

To compare different cut-off values of fasting and random plasma glucose as a screening test for diagnosis of gestational Diabetes in comparison to the 50 grams Glucose Challenge Test [GCT]. Comparative, cross-sectional study. This study was carried out between July 2006 to September 2007 at Departments of Pathology, Obstetrics/Gynaecology and Medicine, PNS Rahat Hospital, Karachi. A total of 53 pregnant subjects at 24-28 weeks of pregnancy were selected to undergo random and fasting blood sugar level estimation and 50-g GCT. All the subjects later underwent 100-g OGTT as well. The results were evaluated by both "Carpenter and Coustan criteria" and "NDDG criteria". The results of random plasma glucose random [cut-off >/= 11.1 mmol/L], fasting plasma glucose [cut-off > 5.3 mmol/L and > 5.1 mmol/L] and plasma glucose results post 50-g GCT [cut-off >/= 7.8 mmol/L and >/= 7.2 mmol/L] were evaluated against 100-g OGTT results through ROC curve analysis. Finally, various diagnostic parameters including sensitivity, specificity, predictive values, likelihood ratios [LR] and efficiency were evaluated. Nineteen subjects were diagnosed to have GDM as per the "Carpenter and Coustan criteria" and 13 met the "NDDG criteria" as per the results of 100-g OGTT. Fasting plasma glucose at was the most efficient investigation at cut- off of 5.1 mmol/L sensitivity=66.66%, specificity=81.25%, PPV=70%, NPV=78.78%, LR+ =3.56, LR- =0.41, efficiency=75.47%. At the cut-off value of 5.3 mmol/L, the results had 64% sensitivity, 85.71% specificity, 80% PPV, 72.72% NPV, 4.48 LR+, 0.42 LR-, 75.97% efficiency]. It was followed by plasma glucose post 50-g GCT [53.57% sensitivity at cut-off of >/= 7.2 mmol/L and 54.54% sensitivity at cut-off of >/= 7.8 mmol/L]. Fasting plasma glucose is a better investigation for the screening of gestational Diabetes than plasma glucose post 50-g glucose challenge