The Application of Six Sigma Techniques in the Evaluation of Enzyme Measurement Procedures in China

Recently, Six Sigma techniques have been adopted by clinical laboratories to evaluate laboratory performance. Measurement procedures in laboratories can be categorized as "excellent", "good", and "improvement needed" based on sigma (σ) metrics of σ ≥ 6, 3 ≤ σ < 6, and σ < 3, respectively. The quality goal index (QGI) was further investigated for measurement procedures with σ ≤ 3. Improvements of the procedures were recommended based on QGI: QGI < 0.8 indicates that the precision of the procedure needs to be improved; QGI > 1.2 indicates that the trueness of the procedure needs to be improved; 0.8 ≤ QGI ≤ 1.2 indicates that both the precision and trueness of the procedure need to be improved.

Fresh frozen sera containing seven enzymes (ALT, ALP, AMY, AST, CK, GGT, and LDH) were sent to 78 clinical laboratories in China. The biases for measurement procedures in each laboratory (Bias) were calculated based on the target values assigned by 18 laboratories performing IFCC (International Federation of Clinical Chemistry and Laboratory medicine) recommended reference methods. The imprecision of each measurement procedure was represented by coefficient variations (CV) calculated based on internal quality control (IQC) data. The σ and QGI values were calculated as follows: σ = (TEa-Bias)/CV; QGI = Bias/(1.5 x CV). TEa is allowable total error for each enzyme derived from biological variation.

Our study indicated that 7.9% (6/76, ALP) to 31.0% (18/58, AMY) of the participating laboratories were scored as "excellent" (σ ≥ 6), 21.1% (16/76, ALP) to 41.3% (31/75, CK) of the laboratories were scored as "good" (3 ≤ σ < 6), and 31.0% (18/58, AMY) to 71.1% (54/76, ALP) of the laboratories need to improve their enzyme measurement procedures (σ < 3). For those with σ < 3, QGIs were further calculated. Based on QGI values, 8.6% (5/58, AMY) to 35.9% (28/78, LDH) of the laboratories (QGI < 0.8) need to improve the precision of the procedures, 8.0% (6/75, CK) to 52.6% (40/76, ALP) of the laboratories (QGI ≤ 1.2) need to improve the trueness of the procedures; and 2.7% (2/75, AST) to 16.3% (8/49, GGT) of the laboratories (0.8 ≤ QGI ≤ 1.2) need to improve both the precision and trueness of the procedures.

Even though rapid progress has been made to standardize serum enzyme measurements in China in recent years, our study using Six Sigma techniques still suggested that approximately 31.1% to 71.0% of the laboratories need to improve their enzyme measurement procedures, either in terms of precision, trueness, or both.