Performance evaluation of an automated method for accurate identification of glucose-6-phosphate dehydrogenase deficiency on the BS-2800M analyzer

The measurement of glucose-6-phosphate dehydrogenase (G6PD) activity was currently used laboratory method for the diagnosis of G6PD deficiency (G6PDd). This study evaluated the analytical and diagnostic performance of a fully-automated method for the one-time quantitative measurement of G6PD activity normalized per unit of hemoglobin (G6PDH) on the Mindray BS-2800M analyzer, using 1561 whole blood samples. For analytical performance, the G6PDH (U/g Hb) showed more stable repeatability performance with coefficient of variation (CV) of 4.48% ± 3.86%, compared with the relatively large CVs in G6PD and Hb detection alone. G6PDH exhibited linearity from 0.37 to 20.86 U/g Hb, which could almost cover the known very low level and very high level, with a carryover rate of less than 1%. The minimum applicable level of Hb for blood sample was at 200 g/L, in which the G6PDH automated method results were not be interfered significantly compared with manual method. Sample stability studies indicated that G6PDH results remained stable for up to 4 h at room temperature and 48 h at 4 °C, but were significantly reduced after storage at - 20 °C (P < 0.05). To evaluate the diagnostic performance of G6PDH, the reference intervals in 806 non-neonates were established as < 2.31 U/g Hb for deficient males/females, 2.31-6.17 U/g Hb for intermediate females. For neonates, they were < 4.14 and 4.14-11.03 U/g Hb respectively based on 122 samples. Comparison of genetic results from 185 samples revealed that G6PDH achieved accuracy of 91.35%, surpassing conventional G6PD single enzymatic method (77.30%) and G6PD/6PGD (82.70%). The kappa coefficient and chi-square analysis indicated that G6PDH phenotypic classifications and genetic results had significant correlation. Furthermore, in 64 of heterozygous females, G6PDH had higher positively detection rate (87.50%) than G6PD single enzymatic method (39.06%) and G6PD/6PGD (51.56%). In summary, the automated G6PDH method offered a valuable approach for clinical laboratory, as it not only had favorable analytical performance for improving testing efficiency, but also provided reliable diagnostic performance for G6PDd, especially holding significant value in precise screening female heterozygotes, early prevention of G6PDd and guiding safe clinical medication.