Objective To evaluate the clinical application of Hb-A2 detection of thalassemia including-alpha-thalassemia and beta-thalassemia. Methods There were 70 Southeast Asia type alpha-thalassemia heterozygosity and 43 normal gene type resulted from Gap-PCR, and 33 beta-thalassemia heterozygosity and 36 normal gene type diagnosed by reverse dot blot (RDB). Cutoff value of Hb-A2 in finding of alpha-thalassemia and beta-thalassemia was present.The sensitivity, specificity, positive and negative likelihood ratio, and receiver operator characteristic curve (ROC) were used to evaluate the significance of Hb-A2 level in screening thalassemia in adults. Results The borderline values between normal and pathological range of Hb-A2 were ≤2.6% in alpha-thalassemia and ≥3.8% available in beta-thalassemia. Sensitivity and specificity were 59% and 54% in alpha-thalassemia, 76% and 83% in beta-thalassemia, respectively. ROC showed that the area covered beneath the curve of beta-thalassemia was more than that of alpha-thalassemia. Conclusions The cutoff between normal and pathological range of Hb-A2 is ≤2.3% in alpha-thalassemia and ≥3.8% in beta-thalassemia. Hb-A2 detection has higher sensitive and specific in beta-thalassemia diagnosis, but lower in alpha-thalassemia diagnosis.
Objective To determine the best threshold value of hemoglobin A2 (HbA2) for diagnosis of β-thalassemia (β-thal) carriers by using high performance liquid chromatography (HPLC), and to improve the application value of HbA2 as a diagnostic index for β-thal carriers to reduce the rates of missed diagnosis and misdiagnosis. Methods Using reverse dot blot (RDB) as a gold standard method, HbA2 results of 1 007 β-thal carriers and 606 normal controls in the past two years determined by HPLC were divided into true positive, false positive, true negative, and false negative based on the different threshold values of HbA2 results. Then, the evaluation indexes such as sensitivity, specificity, positive and negative likelihood ratio, and Youden’s index were evaluated. Next, the receiver operator characteristic (ROC) curve was drawn to determine the best threshold value of HbA2 for diagnosis of β-thal carriers by HPLC. Results If ≥4.0% was taken as the threshold value of HbA2 for diagnosis of β-thal carriers by HPLC, the evaluation indexes values were shown as follows: sensitivity 99.21%, specificity 99.34%, positive likelihood ratio 150.30, negative likelihood ratio 0.008, and Youden’s index 0.99. The Youden’s index was better than the other threshold values, and the corresponding tangent point was the peak point of the ROC curve. Conclusion When ≥4.0% serves as the best threshold value of HbA2 for diagnosis of β-thal carriers using HPLC, integrated evaluation performance of the corresponding sensitivity and specificity is the most ideal, and the authenticity of the diagnostic test is the best.
ObjectiveTo observe and analyze the effect of HbA1c level on macular microcirculation in patients with type 2 diabetes mellitus (T2DM).MethodsA cross-sectional study. One hundred and twenty-four T2DM patients (124 eyes) without diabetic retinopathy who diagnosed by the examination of fundus color photography in Lixiang Eye Hospital Of Soochow University during September to December 2017 were enrolled in this study. There were 59 males (59 eyes) and 65 females (65 eyes), with the mean age of 65.06±7.99 years old. All patients underwent BCVA, fundus color photography, and OCT angiography (OCTA). The history of diabetes, hypertension and dyslipidemia were recorded in detail. According to the HbA1c level, patients were divided into three groups, HbA1c ideal control group (group A, HbA1c ＜7%, 67 eyes), HbA1c control group (group B, 7%≤HbA1c≤9%, 44 eyes), and HbA1c poor control group (group C, HbA1c＞9%, 13 eyes), respectively. The 3 mm×3 mm range of the macular area was scanned by OCTA instrument. The vascular density (VD) and skeleton density (SD) of nonsegmented retinal layer (NRL), superficial retinal layer (SRL) and deep retinal layer (DRL) in the macular area and foveal avascular zone (FAZ) area, non-circularity index, axial rate (AR) of SRL were measured. The correlation between HbA1c, BCVA and VD, SD of NRL, SRL, DRL was analyzed statistically with Spearman correlation test. The correlation between systemic factors and the above indicators was analyzed statistically with linear regression analysis.ResultsThe results of linear regression analysis showed that HbA1c was significantly correlated with VD (t=−3.237, −3.156, −2.050) and SD (t=−0.3.45, −3.034, −2.248) of NRL, SRL and DRL (P＜0.05); but no correlation with FAZ, non-circularity index and AR (t=1.739, 0.429, 1.155; P＞0.05). The differences of VD (F=6.349, 5.981, 3.709), SD (F=7.275, 6.085, 1.904) and AR (F=0.027) of NRL, SRL and DRL in group A, B and C were statistically significant (P＜0.05); but the differences of FAZ (F=1.904), non-circularity index (F=0.280) was not statistically significant (P＞0.05). Significant differences (P＜0.05) of VD and SD of NRL were found between group A and B (t=1.987, 2.201), group A and C (t=3.365, 3.572), group B and C (t=2.010, 2.076). Significant differences (P＜0.05) of VD and SD of SRL were found between group A and B (t=2.087, 2.168), group A and C (t=3.197, 3.194). There were significant differences (P＜0.05) in SD of DRL between group A and B (t=2.239), group A and C (t=−2.519). There was significant difference in VD of DRL between group A and C (t=2.363). The results of Spearman correlation analysis showed that HbA1c was negatively correlated with VD (r=−0.273, −0.255, −0.222; P=0.002, 0.004, 0.013) and SD (r=−0.275, −0.236, −0.254; P＜0.05) of NRL, SRL, DRL; positively correlated with FAZ and BCVA (r=0.221, 0.183; P＜0.05). BCVA was negatively correlated with VD (r=−0.210, −0.190, −0.245) and SD (r=−0.239, −0.207, −0.296) of NRL, SRL, and DRL (P＜0.05), but not correlated with FAZ (r=0.099, P＞0.05).ConclusionThe decrease of macular perfusion and the morphological change of FAZ accompanied by HbA1c increased.