AZ West (Sint Augustinus Veurne)
The measurement of serum 17β-Estradiol is essential to understand physiology, development and health of reproductive processes in both men and women. The determination of estradiol concentration in serum is a major contributing factor in the follow-up and treatment of infertility problems in women. The laboratory of az West currently determines estradiol on the Abbott Architect i1000SR immunoassay analyzer. Other fertility hormones are determined on the Beckman & Coulter Unicel DxI 800 analyzer.
In the past the laboratory direction already tried to consolidate all fertility hormone determinations on one immunoassay platform. Using two analyzers implies loss of time as the primary sample has to be moved from one analyzer to another, and it also necessitates quality control measurements and daily maintenance on an extra analyzer during weekends. Furthermore, the Abbott estradiol assay has limited linearity which makes frequent extra determinations on diluted specimens necessary. However, former estradiol kits from Beckman & Coulter proved not to be sensitive and accurate enough for assessment of estradiol (E2) in low concentrations, which also was obvious from external quality control programs of the Belgian government in the past.
The aim of this work is to evaluate this new Access Estradiol Assay on the Beckman & Coulter DxI 800, which Beckman & Coulter themselves indicate as “sensitive”. For the evaluation of the new testkit, a few performance features were evaluated: sensitivity (LoB, LoQ and LoD), precision (reproducibility and repeatability), total analytical error, uncertainty of measurement, method comparison, measurement range and linearity. Stability in refrigerator and in freezer was also determined.
In conclusion, the new method shows a very good total repeatability and is acceptable within the maximum admissible error as obtained from Sciensano. It also showed a good linearity over a wide range of dilutions. From the LoD value it can be expected that the assay offers improved measurement of low levels of estradiol.
Urine may be a waste product, but it still contains a lot of information and says a bit more about what's going on in the body. Current methodology for the diagnosis of diseases in the urinary system includes microscopic examination of the urine formed elements and chemistry analysis. These methods are time-consuming. The FUS-2000 was purchased to solve this problem. The device is an in vitro diagnostic instrument, which enables chemistry analysis as well as urine formed element analysis and counting through one sampling. So the purpose is to validate the FUS-2000 and consider if the device is able to take over the current routine.
For the validation of the device, a few performance features were evaluated. These features are subdivided into precision (repeatability and reproducibility), trueness and method comparison.
For the evaluation of the repeatability, the positive control of the sediment was measured 10 times one after another. In order to assess the repeatability, the coefficient of variation was calculated. The results showed good test repeatability, with a coefficient of variation (CV) lower than the proposed 8%. This means that the repeatability is acceptable.
For the evaluation of the reproducibility and the trueness, the positive control of the sediment was measured on ten different days. As in the determination of repeatability, the coefficient of variation is calculated. The results of the reproducibility obtained a coefficient of variation also lower than 8% and ensures that the reproducibility is acceptable. The coefficient of variation for the trueness is beyond the limit of -8 and 8%.
This is because the positive control has no reference value, but limit values in which the results should fall. Therefore, for the determination of the trueness, the average of these limit values was taken. So if there is no exact reference value, the accuracy cannot be properly determined.
For the method comparison, the kappa coefficient was calculated. Kappa is a value to determine the agreement between methods and is developed to account the possibility that raters would guess. Like other correlation statistics, the kappa coefficient can range from −1 to +1. A kappa of 0 means that the amount of agreement can be expected from random chance and a kappa value of 1 means perfect agreement between both raters. A concordance table of results is first drawn up before kappa is calculated.
When the FUS-2000 is compared to the microscopic method, the kappa generally showed a good agreement between both methods for white and red blood cells.
For the chemistry analysis, the FUS-2000 was compared to the MIDITRON® JUNIOR II. There was generally a good agreement between both methods. For the ketones a low kappa coefficient was obtained. This is because there are too many negative results and they appear all in the same category in the correlation table. That’s why the chance agreement is high. To solve this problem, more urine samples must be analyzed and compared to each other so the results are more dispersed.
As a general decision, it can be said that the FUS-2000 is useful for the diagnosis of diseases in the urinary system and is suitable for taking over the current daily routine.
Nathalie De Wever