CALIPER Program, Division of Clinical Biochemistry, Pediatric Laboratory Medicine, The Hospital for Sick Children, Toronto, ON, Canada; International UNESCO Center for Health-Related Basic Sciences and Human Nutrition, Mashhad University of Medical Sciences, Mashhad, Iran.
CALIPER Program, Division of Clinical Biochemistry, Pediatric Laboratory Medicine, The Hospital for Sick Children, Toronto, ON, Canada; Department of Laboratory Medicine & Pathobiology, University of Toronto, Toronto, ON, Canada.
Clin Biochem. 2023 Jun;116:79-86. doi: 10.1016/j.clinbiochem.2023.04.001. Epub 2023 Apr 6.
Indirect methods for reference interval (RI) establishment apply statistical techniques to generate RIs for test result interpretation using stored laboratory data. They present unique advantages relative to traditional direct approaches such as fewer resource requirements; however, there is debate regarding their performance. Herein, we aimed to compare indirect and direct approaches for RI establishment by harnessing data from the Isfahan Cohort Study (ICS). This cohort includes both healthy individuals and those with a history of disease, enabling a direct comparison.
Participants were recruited as part of ICS, including 6504 adults aged 34 years and older. Sociodemographic characteristics, anthropometry, blood pressure, various biochemical indices, and hematology parameters were collected. The refineR method was used to establish indirect RIs (before applying exclusion criteria). Direct RIs were calculated using nonparametric methods per CLSI EP28-A3 guidelines (after applying exclusion criteria). Bias ratios were calculated for each parameter to assess significant differences in estimations.
Direct and indirect RI estimations for most hematological and biochemical parameters were comparable. Statistically significant bias ratios between methods were observed for the upper limits of total cholesterol, triglycerides, high-density lipoprotein cholesterol (HDL-C), hemoglobin (female), and platelet count as well as the lower limits of mean corpuscular hemoglobin (female), mean corpuscular volume, hemoglobin, and hematocrit (female).
Data presented indicate RIs derived from direct and indirect approaches are similar, but not identical. Further work should focus on the clinical significance of such differences as well as the investigation of necessary data-cleaning criteria before indirect method application.
间接参考区间（RI）建立方法应用统计学技术，使用存储的实验室数据为检验结果解释生成 RI。与传统的直接方法相比，它们具有独特的优势，例如资源需求较少；然而，关于它们的性能存在争议。在此，我们旨在利用伊斯法罕队列研究（ICS）的数据来比较 RI 建立的间接和直接方法。该队列包括健康个体和有病史的个体，可直接进行比较。
ICS 招募了 6504 名年龄在 34 岁及以上的成年人作为参与者。收集了社会人口统计学特征、人体测量学、血压、各种生化指标和血液学参数。使用 refineR 方法建立间接 RI（在应用排除标准之前）。根据 CLSI EP28-A3 指南（在应用排除标准之后）使用非参数方法计算直接 RI。为每个参数计算偏差比，以评估估计值的显著差异。
大多数血液学和生化参数的直接和间接 RI 估计值相当。方法之间观察到总胆固醇、甘油三酯、高密度脂蛋白胆固醇（HDL-C）、血红蛋白（女性）和血小板计数的上限以及平均红细胞血红蛋白（女性）、平均红细胞体积、血红蛋白和红细胞压积（女性）的下限存在统计学上显著的偏差比。
所呈现的数据表明，直接和间接方法得出的 RI 相似，但不完全相同。进一步的工作应侧重于此类差异的临床意义，以及在应用间接方法之前调查必要的数据清理标准。
