Application of biological variation and six sigma models to evaluate analytical quality of six HbA_1c analyzers and design quality control strategy

Figures & data

Figure 1. Quality target models. Quality targets are shown in colours for the biological variation model (panel A: optimum, desirable, minimum) and with lines for the six sigma-metrics model (panel B: 2–6 σ). Panel C combines both models. The red star represents the Variant II system.

Table 1. The calculation of CV% and bias% for six HbA_1c analyzers (1: 6.31% (number MM2068), 2: 9.64% (number MM2121)).

Name	Specimen	Target	Actual result	CV (%)	Bias (%)	TE (%)	TEa (%)	Acceptable
Ultra2	1	6.31	6.4	0.87	1.43	3.17	7	YES
	2	9.64	9.6	0.52	0.41	1.45	7	YES
Hb9210	1	6.31	6.3	1.23	0.16	2.62	7	YES
	2	9.64	9.5	0.93	1.45	3.31	7	YES
Variant II	1	6.31	6.2	1.36	1.74	4.46	7	YES
	2	9.64	9.5	1.18	1.45	3.81	7	YES
Variant IITurbo 2.0	1	6.31	6.4	1.83	1.43	5.09	7	YES
	2	9.64	9.6	1.5	0.41	3.41	7	YES
Capillarys2FP	1	6.31	6.1	1.43	3.33	6.19	7	YES
	2	9.64	9.4	1.29	2.49	5.07	7	YES
PPI	1	6.31	6.2	2.37	1.74	6.48	7	YES
	2	9.64	9.8	1.91	1.66	5.48	7	YES

Figure 2. Models applied for five instruments (high and low concentrations).

Table 2. σ Values, quality control rules design and quality improvement measures for six HbA_1c analyzers (1: 6.31% (number MM2068), 2: 9.64% (number MM2121)).

Name	Specimen	σ value	QGI	QC rules	N/R	Recommend
Ultra2	1	6.4	1.1	1 − 5s	2/1	–
	2	12.67	0.53	1 − 5s	2/2	–
Hb9210	1	5.56	0.09	1 − 3.5s	2/3	Accuracy
	2	5.97	1.04	1 − 3.5s	2/4	Precision
Variant II	1	3.87	0.85	1 − 3s/2 − 2s/R − 4s/4 − 1s/8 − x	4/2 or 2/4	Precision
	2	4.7	0.82	1 − 3s/2 − 2s	4/1 or 2/2	Precision
Variant IITurbo 2.0	1	3.04	0.52	1 − 3s/2 − 2s/R − 4s/4 − 1s/8 − x	4/2 or 2/4	Accuracy
	2	4.39	0.18	1 − 3s/2 − 2s	4/1 or 2/2	Accuracy
capillarys2FP	1	2.57	1.55	1 − 3s/2 − 2s/R − 4s/4 − 1s/8 − x	4/2 or 2/4	Precision
	2	3.5	1.29	1 − 3s/2 − 2s/R − 4s/4 − 1s/8 − x	4/2 or 2/4	Precision
PPI	1	2.22	0.49	1 − 3s/2 − 2s/R − 4s/4 − 1s/8 − x	4/2 or 2/4	Accuracy
	2	2.8	0.58	1 − 3s/2 − 2s/R − 4s/4 − 1s/8 − x	4/2 or 2/4	Accuracy

Availability of data and materials

The datasets used during the present study are available from the corresponding author upon reasonable request.