Comparison of Oral Glucose Tolerance Test Plasma Glucose Concentrations from Sodium Fluoride and Citrate Buffer Tubes: Does Preservation with Citrate Make a Difference?

Abstract

Background The diagnosis of gestational diabetes Mellitus (GDM) is centred on accurate measurement of plasma glucose obtained from an oral glucose tolerance test (OGTT). However, the measurement of plasma glucose is fraught by many challenges. To minimise ongoing in-vitro plasma glucose loss, the World Health Organisation (WHO) recommends that samples for plasma glucose (PG) should either be measured immediately after collection or collected into tubes containing a glycolytic inhibitor, then placed on ice slurry and separated within 30 minutes of phlebotomy. This recommendation is impractical is resource limited countries like South Africa. Citrate has been proposed as a better antiglycolytic inhibitor with some studies reporting an increase in GDM prevalence when these tubes are used over sodium fluoride (NaF) tubes. Objectives: To compare plasma glucose concentrations obtained from paired NaF and citrate buffer tubes collected at different time points during an oral glucose tolerance test (OGTT) following routine practice at our laboratory and according to World Health Organisation guideline. The second part of the study evaluated the clinical interpretation of plasma glucose results when using citrate buffer tubes over NaF. Methods A prospective observational study was conducted in two cohorts of pregnant women (>18 years) booked for OGTT. In cohort 1, paired NaF and citrate samples were analysed after delayed transport to the laboratory, representing routine practice. In cohort 2, samples were processed according to WHO guidelines. Plasma glucose levels were compared using Passing-Bablok regression and Bland Altman analysis. Cohen’s kappa was used to assess agreement in gestational diabetes mellitus (GDM) classification between tube types. 9 Results Fifty-eight and 51 patients were recruited for cohorts 1 and 2, respectively. Good correlations were observed between NaF and citrate glucose values in both cohorts at all three OGTT time points (Cohort 1: r = 0.941-0.992; Cohort 2: r = 0.982−0993; all p < 0.001). Mean percentage differences (MPD) for Cohort 1 were -8.1%, -4.9%, and -4.5% at fasting, 1h, and 2 h respectively. In Cohort 2, the MPDs were -2.9% (fasting), and -2.5% (both 1h and 2h). All MPDs exceeded the desirable allowable bias specification of 2.3 %. The proportion of women diagnosed with GDM was lower when blood samples were collected using NaF: Cohort 1 NaF identified 24.1% vs 34.4% for citrate tubes and in cohort 2, NaF identified 21.6% vs 25.5.%. Conclusion NaF was less effective at preserving glucose concentrations compared with citrate tubes in both cohorts, with the difference becoming more pronounced when sample processing was delayed. Citrate tubes preserved glucose concentrations without ice, suggesting that citrate inhibitor as better preservative. Accurate GDM diagnosis critically depends on the choice of tube preservative.