Floating mean method for blood cell counter

The moving average method is an indoor quality control method proposed by Bull [1] without quality control products. By observing the mean changes of red blood cell indices (MCV, MCH, MCHC), it can better reflect the status of instruments and reagents. There have been some reports on the application of this method to indoor quality control of blood analyzers, urine sediment analyzers, coagulation analyzers, biochemical analyzers, K+/Na+/Cl- electrodes, etc. [2~6]. It can not only be used to retrospectively evaluate the stability, accuracy and precision of instruments, but also to check the quality of reagents. This paper discusses the indoor quality control of RBC, Hb, MCH, MCHC, MCV, WBC, and Plt by combining the whole blood quality control products. The method and results are reported as follows.
1 Materials and methods
1.1 Experimental materials Sysmex SE-9000 fully automatic blood cell counter and supporting reagents. The instrument was regularly debugged with supporting calibrants, and Bio-Rad BC-12 whole blood quality control (batch number 76521) was used as the quality control product for daily testing.
1.2 Experimental methods The
test results of seven parameters, including RBC, Hb, MCH, MCHC, MCV, WBC, and Plt, of all routine blood specimens (a total of 8268 specimens, with a minimum of 93 and a maximum of 438 specimens per day) on April 30, 2006 were found from the computer records of the Sysmex SE-9000 fully automatic blood cell counter. The mean and standard deviation were calculated, and outliers exceeding ±3SD (total standard deviation) were eliminated according to statistical rules. Then the daily mean (xi) was counted separately, and the total mean (x), total standard deviation (SD) and total coefficient of variation (TCV) were calculated using 30 daily mean values, and the total mean (x) was used as the target value for analysis, and the ±2SD range was used as the standard for the XB analysis method. At the same time, a comparative analysis was performed with Bio-Rad BC-12 whole blood quality control product.

2 Results
The daily mean (xi) of RBC, Hb, MCH, MCHC, MCV, WBC, and Plt of patient samples for 30 days, as well as their total mean (x), total standard deviation (SD), and total coefficient of variation (TCV) are shown in Table 1. The 30-day whole blood quality control test data are shown in Table 2. The changes in the quality control charts using the XB analysis method and the whole blood quality control method within 30 days are shown in Figures 1 to 7, respectively.

3 Discussion
The moving average method is a method designed by Bull et al. in the 1970s for indoor quality control in hematology. It is also called the Bull algorithm or the XB analysis method. The principle is that in the daily blood samples of patients (except for patients with hematological diseases), the mean values ​​of MCV, MCH and MCHC should only vary within a very small range. Therefore, while measuring the samples every day, every 20 samples are analyzed as a batch, and the instrument automatically calculates the XB analysis value of the three parameters of red blood cells. Through the XB analysis method (observing whether the average value is within the controlled range), the working status of the instrument and the quality of the reagents can be observed while testing the patient samples, and problems can be discovered and solved in time. Routine blood analysis is the main basis for diagnosing and distinguishing various blood diseases, infectious diseases, etc., and is the most commonly used and important basic means in clinical medical examinations. According to statistical analysis, the mean values ​​of RBC, Hb, MCH, MCHC, MCV, WBC, and Plt in the blood routine of patient samples every day are relatively constant. As shown in Table 1, the daily average values ​​of the 30-day patient test results varied within a small range, and the TCVs were RBC: 5.19%, Hb: 4.84%, MCV: 0.73%, MCH: 1.84%, MCHC: 1.98%, WBC4.67%, and Plt: 5.11%. As can be seen from Figures 1 to 7, the whole blood quality control curves conform to the Westgard indoor quality control judgment rules for blood routine. The change trends of the RBC, Hb, MCH, MCHC, MCV, WBC, and Plt quality control curves analyzed by the XB analysis method are basically consistent with the whole blood quality control curves. Among them, the average values ​​of MCH and MCHC on the third day exceeded +3SD, and the whole blood quality control products were under control, which may be because the RBC channel was slightly blocked. The whole blood quality control products were processed, and the viscosity was lower than that of fresh whole blood samples, so they were easy to pass through, and the effect on the experimental results was not obvious, so they were still under control; while the whole blood samples of patients had relatively high viscosity, and the red blood cell passage rate was reduced, resulting in an increase in MCH and MCHC, so they were out of control. After clearing and cleaning, the mean values ​​are all within the control range. It can be seen that the whole blood quality control products are treated with stabilizers, preservatives, etc., which makes the quality control data lack a certain comparability with the clinical data. The XB analysis method directly collects quality control data from the measurement results of patient specimens, which better reflects the actual situation of blood analysis than whole blood quality control products. In addition, the better whole blood quality control products are generally imported, with high costs, short validity periods, and unstable properties;
the XB analysis method quality control using patient specimen results can overcome these shortcomings in blood analysis quality control, without increasing quality control costs, and can achieve better quality control results. [page]

In summary, the XB analysis method has good error detection capability and can be used as a supplement to the quality control of whole blood. It can retrospectively evaluate the stability, accuracy, precision and reagent quality of the instrument on the same day, and is especially suitable for comprehensive hospitals with large specimen volumes.
References
[1] Bull BS. A Silldy of various estimators for the derivation of quality control procedures from patient erythro-cyte indices [J]. Am J Clin Pathol, 1974, 61(4): 473-475.
[2] Li Yongxiang, Yang Jiancun. Study on the application of floating mean method of whole blood sample of patients for quality control of blood cell counter [J]. Jiangxi Medical Laboratory, 2003, 21(6): 492-496.
[3] Yue Xin. Quality control of the test results of UF-50 urine sediment analyzer [J]. Chinese Journal of Clinical Medical Research, 2005, 10(1): 28-29.
[4] Huang Jinlian. Application of floating mean method of patient plasma in quality control of PT, APTT and Fib [J]. Journal of Practical Medical Technology, 2003, 10(11): 1247-1248.
[5] John A. Use of medians and average of normals of patients date for assessment of long-term analyticalatability[J].Clin Chem,1996,42(6):888-892.
[6] Fang Yusen, Huang Xudong. Extended use of Easylyte plus K+/Na+/Cl- analyzer electrodes[J]. Journal of Practical Medical Technology,2002,9(1):9-10."#$!