Design of internal quality control program for specific protein testing based on patient risk / 中华检验医学杂志

ABSTRACT

Objective:Patient risk-based statistical quality control (SQC) program was designed for 9 specific protein projects using Westgard sigma rules with run length.Methods:The cumulative coefficient of variation of immunoglobulin (Ig)G, IgA, IgM, C3, C4, rheumatoid factor (RF), antistreptolysin O (ASO), transferrin (TRF) and prealbumin (PA) from the laboratory department of Beijing Tongren Hospital between December 2018 to May 2019 were used as the estimated value of imprecision. The mean of the absolute value of the percentage difference of 10 batches in the laboratory, which was derived from the results of participating the external quality assessment (EQA), was used as the estimated value of bias. The National Center for Clinical Laboratories EQA evaluation criteria was used as an allowable total error (TEa), and the sigma value of each project (σ) was calculated. Westgard Sigma rule with run length was used to design appropriate SQC program for each project, including quality control rules, number of control measurements (N) and frequency of quality control.Results:The sigma value was larger than 6 for SQC procedure of IgG, IgA, IgM, C4 and TRF. SQC could be established with the use of 1 3s rule, number of control measurements (N)=2, number of runs (R)=1, and a run length of 1 000 patient samples. Combined with the average daily workload, internal quality control could be conducted once every 10 days for IgG, IgA, IgM and C4, every 50 days for TRF. The σ was 5.86 for C3, SQC program could be established with run length of 450 using 1 3S/2 2S/R 4s rule (N=2, R=1), combined with average daily workload, internal quality control could be conducted every 4.5 days. σ was between 3 and 4 for RF, ASO and PA. With the use of 1 3S/2 2S/R 4s/4 1s/6 X rule (N=6, R=1), SQC program with a run length of 45 and higher frequency internal quality control activities. Conclusion:It is feasible to use Westgard sigma rules with run length for the laboratories design of personalized risk-based SQC procedures, the method is very simple and intuitive. This tool is valued to be recommended to be actively applied by all clinical laboratories.