[Consistency evaluation of 2 methods in detecting serum insulin-like growth factorⅠ in children]

Objective: To evaluate the consistency of mass spectrometry (MS) and chemiluminescence immunoassay (CLIA) in detecting serum insulin-like growth factor-1 (IGF-1) and IGF-1 standard deviation score (SDS). Methods: This cross-sectional parallel control study prospectively collected the serum samples of 115 children with short stature disorders who were admitted in the Department of Endocrinology, Beijing Children's Hospital, Capital Medical University from February 2020 to December 2021. The serum IGF-1 level was detected by CLIA and MS, and converted to SDS for consistency analysis. Pearson analysis was used to analyze the correlation between the 2 methods, and Deming regression equation was established. Bland-Altman diagram and weighted Kappa coefficient were used to evaluate the consistency of the 2 methods. Results: There were 46 boys (40.0%) and 69 girls (60.0%), aged (8±3) years. Among the 115 cases, 37 were Turner syndrome, 59 were small for gestational age (SGA) at term, 1 was growth hormone deficiency (GHD) and 18 were other diseases. Pearson correlation analysis showed a preferable correlation between IGF-1 measured by the 2 detection methods (r=0.94, P<0.01), and IGF-1 SDS was also significantly correlated (r=0.92, P<0.01). Bland-Altman analysis showed that the consistency of serum IGF-1 levels detected by the 2 methods was poor, and the mean difference between CLIA and MS was 33.38 μg/L. The result detected by CLIA was significantly higher than that by MS, with SDS of 43.51 μg/L (95%CI -51.89-118.7 μg/L). After converting the results to SDS and removing 3 outliers (including 1 GHD patient), the weighted Kappa showed acceptable consistency (κ=0.68). Conclusion: In clinical application, after converting to IGF-1 SDS, IGF-1 detected by MS and CLIA can be used for cross-reference, but too high or too low levels should be cautious about.