Clinical evaluation of a modified "Oxford T 4 -by-column" method for serum thyroxine.

Three commerciallyavailablekitsfordetermining serum thyroxine (T4) by column were evaluated. Results with all were less than satisfactory, but a series of changes in the Oxford method resulted in increased reproducibility, substantially improved recovery, and extended linearity of the standard curve. Results compared well with the thyrometabolic status of patients and normal individuals. The 95% confidence limits for 219 “normal” hospital employees were 2.9-6.2 g/100 m. The suggested normal range for 241 clinically euthyroid patients is 2.5-6.0 g/100 ml. The coefficient of correlation between results of the proposed method and of the PBI method was 0.30; for the T4 by the Murphy-Pattee method it was 0.75.