Objective: To evaluate the efficiency of fasting capillary blood glucose (FCG) measurement as compared with fasting venous plasma glucose (FPG) measurement in screening diabetes and pre-diabetes in low-resource rural settings.
Subjects and methods: In 2010, 993 participants were randomly selected from 9 villages in Yunnan province using cluster sampling method. Samples for FCG and FPG test were obtained after demographics and physical examination. The oral glucose tolerance test was performed in parallel as gold standard for diagnosis. Diagnostic capacities of the FCG measurement in predicting undiagnosed diabetes and pre-diabetes were assessed. The performance of FCG and FPG tests was compared.
Results: Fifty-seven individuals with undiagnosed diabetes and 145 subjects with pre-diabetes were detected. The concordance between FCG and FPG levels was high (r = 0.75, p < 0.001). The area under the curve (AUC) for FCG test in predicting diabetes was 0.88 [95% confidence interval (CI) 0.82-0.93] with the optimal cutoff value of 5.65 mmol/l, sensitivity of 84.2%, and specificity of 79.3%. The corresponding values in FPG tests were 0.92 (95% CI 0.88-0.97) (AUC), 6.51 mmol/l (optimal cutoff point), 82.5% (sensitivity) and 98.3% (specificity), respectively. No significant difference was found in the AUC for the two screening strategies.
Conclusion: FCG measurement is considered to be a convenient, practicable screening method in low-resource rural communities with acceptable test properties.