目的： 探讨采用不同稀释介质稀释的血清标本对化学发光免疫法检测血清雌二醇（estradiol,E₂）浓度结果的影响。方法： 应用全自动化学发光免疫法对50例血清标本进行E₂浓度检测,分别使用仪器配套的标准稀释液S₀及3种替代稀释液（低值混合血清、商品E₂/孕酮稀释液、新生牛血清）作为稀释介质,进行1∶1或1∶4的手工稀释后,上机检测;同时每管均设平行管,进行统计分析。结果： 经配对t检验比较,3种替代稀释介质组与标准稀释液S₀组间的稀释后检测值差异均具有统计学意义（ P值分别为5.03×10^-11,1.44×10^-10及2.79×10^-13）。Pearson相关分析显示,新生牛血清组、商品E₂/孕酮稀释液组、低值混合血清组与标准稀释液S₀组之间的相关系数r分别为0.992 3、0.971 3及0.968 0,新生牛血清组与标准稀释液S₀组间的相关性最为显著（r>0.99）。3种替代稀释介质经线性回归模型校正后的检测值与标准稀释液S₀间比较,差异无统计学意义（ P值分别为0.998 3、0.997 9、0.987 7）。采用Bland-Altman方法分别评估新生牛血清、商品E₂/孕酮稀释液及低值混合血清与标准稀释液S₀之间的一致性,结果显示3种替代稀释介质中分别有2%、6%及8%的检测结果超出一致性界限,最大偏差分别为16.8%、44.8%及45.2%,三者中新生牛血清的一致性最佳。结论： 在多种替代稀释介质中,新生牛血清在高浓度E₂稀释中检测结果最佳,且稳定性好、成本较低、可减小异嗜性抗体的影响,是作为化学发光免疫法测定高浓度E₂的最优替代稀释液。

Objective: To evaluate the impact of different diluting media on determination of serum estradiol (E₂) concentration by using chemiluminescent immunoassay. Methods: Fifty serum samples were collected for E₂ detection using chemiluminescent immunoassay by an automatic biochemical analyzer (UniCel DXI 800, Beckman Coulter). Four diluting media were used, including standard diluent S₀ (Beckman Coulter), pooled low-level E₂ serum (PS), E₂/progesterone standard diluent (Roche), and calf serum (CS). Samples were manually diluted into 1∶1 or 1∶4 using the above 4 media with two parallel tubes and detected by UniCel DXI 800. Multiple statistical methods were applied to analyze the data. Results: Students't test revealed that the E₂ concentration measured by diluting media CS, Roche and PS compared with S₀ showed significant differences statistically (P=5.03×10^-11, 1.44×10^-10 and 2.79×10^-13, respectively). Pearson correlation analysis showed that the correlation coefficient r between diluting media S₀ and CS, Roche, PS were 0.992 3, 0.971 3, 0.968 0, respectively. After correction with linear regression model, the concentration measured by diluting media CS, Roche and PS compared with S₀ showed no significant difference statistically (P=0.998 3, 0.997 9, 0.987 7, respectively). Consistence of E₂ concentration measured by diluting media CS, Roche and PS compared with S₀ was evaluated by Bland-Altman method, the results showed that 2%, 2% and 8% of the samples exceeded the consistence limit, respectively. The maximum bias of three assessed diluting media were 16.8%, 44.8% and 45.2%, respectively. CS shows the best consistence. Conclusions: Among the three assessed diluting media, CS shows the best performance and could serve as a replacement of manufacturer's standard diluent with high stability and lower cost.