The use of six sigma methodology in evaluating the analytical process in clinical laboratories

Objective: The main aim is to ensure and maintain the quality of the total testing process in clinical laboratories. The six sigma method is a significant tool used in laboratories for detecting, evaluating and monitoring errors across all processes, especially in the analytical phase. To achieve this, a quality target should be established that allows for an acceptable level of error without compromising patient safety. The sigma metric also helps determine appropriate frequencies for quality control. There is still uncertainty about the quality control frequency in clinical laboratories. In our study, we aimed to determine whether using tight and wide TEa targets in the monitoring of the laboratory process with the six sigma method and obtaining the bias in the six sigma calculation from external quality assessment (EQA) or internal quality control (IQC) data would make a difference in terms of quality control procedures. Materials and Methods: The 12-month IQC data for 2022 of 12 biochemistry and 8 immunoassay tests studied in the Pamukkale University Hospitals Medical Biochemistry Laboratory were obtained retrospectively from the Laboratory Information System LIS, and the DKD data were obtained retrospectively from the BIORAD-EQAS website. For the tight and wide TEa target, suitable ones were selected among CLIA 2024, Turkey, Rilibak, EFLM BV Minimum and Desirable targets. Bias was calculated using two different approaches based on EQA and IQC data. Median values of 12 months were used for annual bias, CV and sigma values. Sigma levels of tests were calculated using the formula Sigma = (%TEa - %Bias) / %CV. Sigma levels obtained from different TEa targets and bias approaches were compared. Quality control procedures appropriate for the sigma level of each test were determined. Quality Goal Indices (QGI) were calculated to identify sources of issues for tests with sigma levels less than 4. Performance of each test was evaluated using method decision charts. Results: Using the bias obtained from EQA data of biochemistry tests, according to the wide TEa target, tests with annual sigma levels over 6 are triglyceride, ALP, LDH and urea; in the 5-6 range for ALT, albumin, glucose, and chloride and in the 4-5 range for creatinine, total cholesterol and total protein tests. According to the wide TEa target, the sigma level for all tests was 4 and above. According to the strict TEa target, tests in the range of 4-5 are ALP, LDH and triglyceride; tests in the range 3-4 are AST, ALT, glucose, T. cholesterol and urea; tests below 3 were determined as T. protein, albumin, chloride and creatinine. Using EQA data, none of the tests achieved a sigma level above 5 relatives to the stringent TEa target. Sigma levels calculated using the wide TEa target with bias derived from IQC data were above 6 for triglyceride, LDH, ALP and urea; in the 5-6 range for AST and T. protein; in the 4-5 range for glucose, chloride, creatinine, albumin, ALT and T. cholesterol. Using IQC data, only the triglyceride test had a sigma level above 5, according to the stringent TEa target. The sigma level was found to be in the range of 4-5 for ALP and LDH, in the range of 3-4 for ALT, AST, T. cholesterol, urea, glucose and below 3 for T. protein, creatinine, chloride and albumin. Choosing the strict TEa target resulted in lower sigma levels compared to the wide TEa target. For immunoassay tests, sigma levels calculated using bias from both EQA and IQC data were above 6 for SHBG, TSH, estradiol and testosterone. Sigma levels of LH and FSH tests in both bias approaches were in the range of 5-6. For prolactin had a sigma level in the 5-6 range when using EQA bias, and 4-5 range when using IQC bias. For fT4 test, sigma levels were in the range of 4-5 with both approaches. Control procedures were recommended for tests with sigma level below 3, 13S/22S/R4S/41S, N:3 R:2; for tests in the range 3-4, 13S/22S/R4S/41S, N:2 R:2; for tests in the range 4-6, 12.5S N:2 R:1 and 13S N:2 R:1 for tests above 6. Conclusion: There was no significant difference in sigma levels between EQA and IQC approaches for bias used in sigma calculation. It has been observed that the main determining factor in sigma values ​​is the selection of a tight or wide TEa target. Each laboratory should determine a TEa target appropriate to its own conditions. Key Words: Quality control, six sigma, analytical process, total allowable error, TEa