QUALITY CONTROL, QC Methods/Application
6-48 Sysmex SF-3000 Operator's Manual -- Revised September 1995
10.3 X M Control
The SF-3000 has X M Control for monitoring instrument stability over time. X M
Control calculates the weighted averages of batches of samples in normal routine
analysis, and compares the control values obtained for the different samples. It cannot
control a batch until it has been processed, but is effective in checking hour-by-hour
changes and can detect changes in the instrument and in the reagents with great
sensitivity. However, X M Control is affected by variations in the analysis values
within a batch, and if a limit error occurs, the sample analysis values should be checked.
The control program uses the X B algorithm published by Dr. Brian S. Bull. This
algorithm is formulated using the three blood cell indices (MCV, MCH and MCHC).
X M Control is used in the SF-3000 because of the large number of control parameters
(28).
NOTE: The "M" of " X M" stands for "moving". (Or, more precisely, "weighted
moving average".)
This program's computations are performed using the following equations:
XM( i ) = XM ( i -1) + SGN { SGN [ X( j , i ) - XM ( i -1) ] X( j , i ) - XM ( i -1) } F
F =
N
j=1
N
SGN X( j , i ) - X
M ( i -1) X( j , i ) - XM ( i -1)
N
j=1
2
{ }
Figure 6-40: Equation (1)
Here, X M (i) is the X M value for the i-th batch, X M (i-1) is the X M value for the
(i-1)th batch, X (j, i) is the j-th analyzed value in the i-th batch, and N is the number of
samples in a batch.
The X M quality control program is based on the fact that once the characteristics of a
population have been established for a particular institute, the parameters will remain
stable within these limits. The characteristics of the population can be represented by the
X M limits, which are computed using at least 30 X M control data. Equation 1
contains the techniques of trimming and smoothing.
To Next Page
Loading...
