A comparison between point-of-care testing and venous glucose determination for the diagnosis of diabetes mellitus 6-12 weeks after gestational diabetes

Aim To evaluate point-of-care-testing (POCT) for the diagnosis of Type 2 diabetes mellitus 6-12 weeks post-partum in women with gestational diabetes (GDM). Methods Post-partum glucose assessment (75-mg oral glucose tolerance test, OGTT) was performed prospectively in 122 women with GDM (1 November 2015 to 1 November 2017) at Tygerberg Hospital, Cape Town, South Africa. Individuals with known pre-existing diabetes were excluded. The accuracy and clinical utility of POCT (capillary finger-prick) were compared with laboratory plasma glucose (hexokinase and glucokinase methods). The OGTT consisted of two time points (fasting and 2 h) during which concurrent glucose samples (POCT and laboratory) were obtained. Bland-Altman plots and paired analysis were used to assess the analytical accuracy of POCT, whereas its diagnostic performance was determined using positive and negative predictive values to calculate specificity and sensitivity. Results Spearman's ranked correlation analysis indicated a strong association between POCT and laboratory glucose values at both OGTT time points (fasting, r = 0.95, P < 0.0001; 2 h, r = 0.88, P < 0.0001). Thirty-six women were diagnosed with Type 2 diabetes based on gold standard laboratory glucose levels (fasting > 7 mmol/l; 2 h > 11.1 mmol/l). POCT correctly identified Type 2 diabetes in 78% of women (28 of 36) with a positive predictive value of 89.3% and a negative predictive value of 96.7% at the fasting time point. The sensitivity and specificity of POCT to diagnose Type 2 diabetes were 89% (fasting), 85.7% (2 h) and 96.7% (fasting), 98.5% (2 h) respectively. POCT proved less sensitive to diagnose pre-diabetes (69%) but displayed satisfactory specificity (92%) at both time points assessed. Conclusion POCT accurately identifies women with Type 2 diabetes 6-12 weeks after GDM.