A new approach to calculating the Sigma Metric in clinical laboratories

COŞKUN A., SERTESER M., Kilercik M., Aksungar F., Unsal I.

ACCREDITATION AND QUALITY ASSURANCE, vol.20, no.2, pp.147-152, 2015 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Editorial Material
  • Volume: 20 Issue: 2
  • Publication Date: 2015
  • Doi Number: 10.1007/s00769-015-1113-8
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.147-152
  • Keywords: Bias, Sigma Metric, Six Sigma, Tolerance limit, z transformation
  • Acibadem Mehmet Ali Aydinlar University Affiliated: Yes


In clinical laboratories, the performance of a process as Sigma Metric (SM) is calculated by the equations derived by Westgard. In the present study, we found that the Westgard equations do not reflect the real performance of the process and that the SM calculated using these equations is lower than the real SM. We measured the substance concentration of ten analytes (glucose, urea, creatinine, cholesterol, calcium, magnesium, phosphorus, LDH, sodium, and potassium) in serum and calculated the SM for each using the Westgard equations and z transformations. The SM values for the same measurand using the Westgard equations based on either the absolute or the relative (percentage) results were not equal to each other, and those related to calcium and sodium were even lower than 0. The SM obtained from the z transformation was higher than that from the Westgard equations, and none were lower than 0. We concluded that the equations suggested by Westgard to calculate the SM do not cover all of the data produced by the process and do not reflect reality. From our research, the z transformation was the optimum method of calculating the actual SM for the process.