Quality control of radioimmunassay of thyroid related hormones using commercial RIA kits

The Passive Quality Control data of 31 batches of T3, 56 batches of T4, 28 batches of TSH, 8 batches of free T3 and 7 batches of free T4 RIA over a duration of one year were collected to estimate the overall reproducibility of results and level of error observed in assays employing Amersham RIA kits. The parameters recorded were: standard and unknown scatter, RER (response error relationship) parameters, standard curve parameters and quality control pool results. Cumulative RER and IP plots were drawn using individual batch data to estimate inter batch variation of error. The data analysis was performed on IBM-compatible computer using programme supplied by IAEA. Overall random error (non-counting statistics %CV) in all replicate counts was less than 7%. Mean RER parameters also suggest an overall low level of random errors in all assays. The individual batch RER's do not cross 10% error limit. This shows that random errors in out assays are not affecting the assay performance. A view of average imprecision profiles show that T3 and T4 RIA's are highly precise. TSH assay is relatively more imprecise in the normal range. FT3-RIA gives acceptable precision. In FT4-RIA the error is greater than 10%. A comparison of the overall between batch scatter (random variation) of curve parameters shows that parameters of T4 RIA's are most reproducible. Parameters of T3, TSH and FT4 RIA's have intermediate reproducibility, whereas FT3-RIA parameters are least reproducible. A comparison of BBCV's of quality Control results shows that T4-RIA is giving most reproducible results. T3, TSH and FT3 RIA's have intermediate reproducibility. FT4-RIA is least reproducible. Assays of each category show a negative drift except FT3-RIA where a positive trend is dominant. The average values of drift are within 10% limit. It is concluded that T4-RIA stands at the top as far as assay performance is concerned. (author)