Evaluation of dried blood spot TSH-IRMA kit produced for the first time in the I. R. Iran

The production of neonatal TSH kits for the measurement of the analyte on filter papers has been considered a necessity for national screening program of congenital hypothyroidism (CH) in the I. R. Iran. This report aims at evaluating the afore mentioned kit, which has been produced for the first time in the I. R. Iran and used by Endocrine Research Center in the screening program for CH. From February 1998 to June 2001, cord dried blood spot (DBS) samples applied onto filter papers were sent to Endocrine Research Center laboratory from 8 hospitals in Tehran and Damavand for TSH measurement using a two-site IRMA kit produced by Atomic Energy Organization of Iran according to the NETRIA company instructions. Validity of the method including precision (intra- and inter-assay tests), accuracy (recovery tests), sensitivity, and specificity were assessed. Also, absolute values in difference of TSH values of 41 randomly selected cord DBS samples measured by both the present and DRG (ELISA) neonatal TSH kits were calculated and their correlation was assessed using Pearson’s correlation coefficient. For statistical analyses, the SPSS version 9.05 software was employed. Cord DBS TSH levels of 20107 samples were measured. Three CH cases with borderline cord TSH values (23-35 mIU/L) were detected. The sensitivity of the method was 1 mIU/L. The coefficients of variation (%) were 10.7, 8.7, 9.4, and 9.6 (intra-assay) and 12.2, 9.1, 10.2, and 11.1 (inter-assay) for TSH values of 3, 23, 45, and 82 mIU/L. Recovery tests showed a recovery of 108, 112, and 109% for the proportion of the measured (13, 36, 69 mIU/L) to the expected (12, 32, 63 mIU/L) TSH concentrations. Specificities of the method for TSH, LH, hCG, FSH, PRL, and GH were 100, <0.005, <0.005, <0.003, <0.01, and <0.02%, respectively. The differences of corresponding TSH values as measured by both kits were less than 1.5 mIU/L in 95% of comparisons. There was a significant correlation between TSH levels measured by both kits (r = 0.96, p < 0.001). The quality of TSH kits was considered acceptable for the detection of CH, especially in mildly abnormal cord DBS values, according to the findings of the screening program for CH and of the methodologic evaluation.