Introduction

The free thyroxin (FT4) is the main hormone secreted by the thyroid gland in the blood serum. The BIOMAG laboratory is a clinical laboratory. It uses an analytical method based on electro-chemiluminescence to measure it.
The Roche Diagnostics company has developed a new method to analyse FT4. The analyses were conducted by an automaton called Modular. This automaton was provided by the company. The laboratory performed a methods comparison. The aim is to demonstrate that two methods are not statistically different.
Therefore we can use the two methods for the same purpose and keep the medical history of the patients.

Experimental conditions

In order to carried out the methods comparison, results obtained with each method are needed. The experience was conducted in the smallest span possible. 30 samples were analysed with the two methods using the same machine. Those samples were distributed so as to cover the working range of the method for the free thyroxin.
Following the collecting of all the data, the difference di was calculated:

d_i= x_i- y_i
The obtained values correspond to :
-     x_i for the method already used in the laboratory.
-     y_i for the new method.
A representation of this difference according to the FT4 concentration values can be found on the graphic 1 (d_i = f (x_i)).

Results

To determine if the new method is not statistically different from the previous one, some thresholds are defined. Those thresholds were determined by the SFBC’s values. It is a French association that provides criteria for the evaluation of the performance of in vitro diagnostic methods. These values are defined for comparison method and specific for each biological test.
The obtained value calculating the difference, for a level of concentration, is described as the ordinate. The reference value, for the same level, is described as the abscissa. An upper (UT) and a lower (LT) thresholds are set up; one with the criteria defined by the SFBC negative and one with them positive.
As we can see on graphic 1, all the obtained data (differences between the two methods) are gathered between the two limits. It means there are no conflicting results for those two methods. The two methods can be considered as not statistically different from each other.

Conclusion

Since the two methods were proved as statistically comparable, the laboratory accepted to use the new method. It is said to be more precise than the previous one. The new method is used to measure free thyroxin.

Difference di according to FT4's concentration

SELARL BIOMAG
Laboratoire de la gare 3,
Avenue Jules Uhry
60100 Creil

SECRETARY
Catherine BEKHALED