Objective: Graves' disease is an autoimmune disease which results in excessive production of thyroid hormones and is the most common cause of hyperthyroidism. TRAb has a key role to confirm the clinical diagnosis of Graves' disease. The different measurement methods have been developed because of the different bioactivity types of these antibodies. Radioimmunoassay (RIA) is the recommended method of measurement in Graves' diagnostic guidelines. Recently, use of immunebased automated methods instead of manually RIA measurement is discussed. In our study, we aimed to compare the availabilities of automated immunoassay and manual RIA methods for TRAb analysis.

Materials and Methods: According to “Patient Based Comparison Methods of Clinical Laboratory Standards Institute”, the serum samples of 55 patients were analyzed by RIA and ECLIA methods. The systematic and random error rates were determined for both methods. The method's validities were evaluated for medical decision criterions.

Results: While total error (TE) of the RIA method was 19.81 (%), TE rate of ECLIA method was 41.6 (%). According to regression analysis, the correlation between two methods was good level (r = 0.8). At the lower levels of TRAb, the differences between RIA and ECLIA were higher than the others. When borderline test results were evaluated as any thyroid disease except Graves'; the negative predictive value of ECLIA method was 97.8 (%), the validity of the test was 94.3 (%).

Conclusion: ECLIA method may be preferable for routine TRAb analysis to diagnosis Graves' disease, although it has high total error rate compared to RIA. It has high diagnostic validity and negative predictive values. It is known that automated methods may have lower random error compared to manual methods. On the other hand, the cost effectiveness of automated ECLIA method should be evaluated because of its requirements such as analyzer and supplies.