Assessment of Analytical Quality in Clinical Laboratory Using Sigma Metrics

Abstract

Background: Analytical quality in clinical laboratories is crucial for generating reliable test results that directly influence diagnosis and patient management. Traditional indicators, such as precision and accuracy, provide only partial assessment. Six sigma metrics offer a comprehensive, quantitative approach by integrating total allowable error (TEa), bias, and imprecision to evaluate the overall performance of analytical methods. Objectives: To assess the analytical performance of routine biochemical analytes using six sigma metrics and classify analytes according to sigma performance, and to identify analytes that require method improvement. Methods: A retrospective observational study was conducted at Biochemistry department DRPGMC, Tanda, Himachal Pradesh, India, using Internal Quality Control data and External Quality Assessment (EQA) results of 6 months from a clinical biochemistry laboratory. Imprecision (coefficient of variation [CV %]) was calculated from daily quality control (QC) data, and bias (%) was derived from EQA peer-group mean values. TEa% values were adopted from the Clinical Laboratory Improvement Amendments (CLIA) guidelines. Sigma metrics were calculated using the formula: σ = TEa ˗ ∣Bias∣ ÷ CV. Analytes were categorized into high (≥6σ), moderate (3–5.9σ), and low (<3σ) performance groups to guide QC rule selection. Results: Sigma metrics varied across analytes and required the TEa criteria applied. When assessed using the CLIA-88 TEa limits, triglycerides and high-density lipoprotein cholesterol (HDL-C) demonstrated high sigma performance (≥6σ), indicating excellent analytical precision. Moderate sigma performance (3–5.9σ) was observed for glucose, uric acid, alanine aminotransferase, aspartate aminotransferase, alkaline phosphatase, total protein, cholesterol (at level 3), and calcium (at level 3), necessitating multi-rule quality control strategies. In contrast, urea, creatinine, albumin, and phosphorus exhibited poor analytical performance with sigma values <3σ, indicating the need for improving the method. However, when sigma metrics were recalculated using the more stringent CLIA-2025 TEa limits, a further decline in analytical performance was observed. Uric acid, liver enzymes, and total protein demonstrated sigma values <3σ under the revised criteria, whereas triglycerides (at level 3) and HDL-C consistently maintained high sigma performance (≥6σ) despite the narrower allowable error limits. Conclusion: Six sigma assessments provided a comprehensive and quantitative measure of analytical quality in a clinical laboratory. Incorporating sigma metrics into routine quality assurance enhances reliability, optimizes QC protocols, and strengthens patient safety.