|
|
Background: The EP6-A protocol and Dr. Kroll’s EP6-A extended methods are newly approved schemes for assessing linearity. Linearity surveys are used to evaluate the performances of clinical laboratories and improve the quality of tests.
Methods: In accordance with the instruments used, the data of blood calcium were divided into four groups: Olympus (70 labs), Hitachi (57 labs), Beckman CX (17 labs), and Beckman LX (28 labs). Statistically significant linearity and nonlinearity were judged using polynomial regression analysis, as recommended by the Clinical and Laboratory Standards Institute EP6-A guideline. The imprecision of analytical data were analyzed and the extents of nonlinearity were determined using Dr. Kroll’s EP6-A extended methods.
Results: In samples analyzed using Olympus, Hitachi, Beckman CX, and Beckman LX instruments, the imprecision ranged from 0.05 % - 11.92 %, 0.16 % - 10.39 %, 0.42 % - 2.73 %, and 0.10 % - 4.10 %, respectively. The imprecision for two laboratories in the Olympus and Hitachi groups exceeded the acceptable limits, while the imprecision for the remaining laboratories were precise enough for a linear survey. The average deviations from linearity ranged from 0 % - 2.74 %, 0 % - 3.92 %, 0 % - 2.25 % and 0 % - 3.73 %, respectively. The linearity with statistical significance was 28.6 %, 43.9 %, 41.2 %, and 3.6 %, respectively. The linearity with clinical significance was 70.0 %, 54.4 %, 58.8 %, and 96.4 %, respectively. The percentages of laboratories whose blood calcium analyses met the linearity criteria were 98.6 %, 98.3 %, 100 %, and 100 %, respectively.
Conclusions: The majority of the laboratories involved in the linearity survey showed good capability for blood calcium testing.
DOI: Clin. Lab. 2011;57:327-331
|
