Discrepancies in Sigma Metrics Driven by Total Allowable Error Variability: Implications for QC Strategy and Laboratory Efficiency.

Beyond the basics: Sigma scores in laboratory medicine with variable total allowable errors (TEa)

BackgroundLaboratories use quality control processes to monitor and evaluate analytical performance in terms of precision and bias. Sigma metrics provide an objective assessment of laboratory quality using the total allowable error as an additional parameter.ObjectiveThis study aimed to determine the sigma metrics of analytes when using different total allowable error guidelines.MethodsA retrospective analysis was performed on 19 general chemistry analytes at Charlotte Maxeke Johannesburg Academic Hospital in South Africa between January 2017 and December 2017. Sigma metrics were calculated on two identical analysers, using internal quality control data and total allowable error guidelines from the Ricos biological variation database and three alternative sources (the Royal College of Pathologists of Australasia, the Clinical Laboratory Improvements Amendment, and the European Federation of Clinical Chemistry and Laboratory Medicine).ResultsThe sigma performance was similar on both analysers but varied based on the guideline used, with the Clinical Laboratory Improvements Amendment guidelines resulting in the best sigma metrics (53% of analytes on one analyser and 46% on the other had acceptable sigma metrics) and the Royal College of Pathologists of Australia guidelines being the most stringent (21% and 23%). Sodium and chloride performed poorly across all guidelines (sigma < 3). There were also month-to-month variations that may result in acceptable sigma despite poor performance during certain months.ConclusionThe sigma varies greatly depending on the total allowable error, but could be a valuable tool to save time and decrease costs in high-volume laboratories. Sigma metrics calculations need to be standardised.

Background: The Six Sigma methodology has become a standard in quality assessment and management, widely adopted in various industries since the mid-1980s. This methodology is applicable wherever the outcome of a process needs measurement. One area where Sigma metrics find utility is in laboratory services, particularly in evaluating the quality assurance of biochemical parameters. Aims: This study aimed to measure and analyze the sigma metrics of thyroid hormones T3, T4, and TSH within the framework of NABL accreditation (National Accreditation Board for Testing and Calibration Laboratories). Materials &amp; Methods: The study was designed as a retrospective observational study, utilizing quality control data from the Clinical Biochemistry section spanning from January 2021 to December 2021 retrieved from laboratory records. The data considered included Bio-Rad Lyphochek Immunoassay Internal Quality Control (IQC) Control and Bio-Rad External Quality Assurance Scheme (EQAS) Immunoassay monthly program data. Calculations for Coefficient of variation (CV %), Bias %, and Total Error allowable (TEa) were made to derive the sigma metric. The formula used for sigma calculation was Sigma Σ(σ) = (TEa% - Bias%) / CV%, while the Quality Goal Index (QGI) ratio was determined using the formula QGI = Bias / 1.5CV. Results: The results categorized the assay performance based on sigma levels: &gt;6 denoted world-class performance, 5-6 indicated excellence, 4-5 represented good performance, 3-4 was deemed acceptable, and 2-3 signified poor performance. All analytes demonstrated sigma levels below 6 (2.61-5.85) varying from poor to excellent and the QGI was below 0.8, indicating a need to enhance the precision of assays. Conclusion: The Six Sigma metric analysis serves as a benchmark tool that can aid laboratories in enhancing assay performance, formulating protocols for Internal Quality Control (IQC), and evaluating existing laboratory procedures. Keywords Six sigma, Quality Goal Index, Internal Quality Control

Background: Application of sigma matrics across clinical laboratory diagnostic process enables identification of errors in methods together with quality goal index can help point out the cause of unacceptable process performance. We believe the analytical performance of the clinical biochemistry laboratory standards do not meet the required quality according to six sigma standards. Aim and Objective: The study aimed at performance evaluation of clinical biochemical laboratory analytical process using sigma metrics and quality goal index. Methods: Sigma (σ) for the 20 analytes performed on HITACHI 7600 - 110 was calculated based on absolute bias, total allowable error (TEa) and coefficient of variation (CV). Two level internal quality control data obtained retrospectively from December 2020 to May 2021 was used to calculate the coefficient of variation. Bias was obtained from six month external quality control data. Total allowable error was obtained from three different sources. Quality goal index (QGI) was calculated to identify the cause of unwanted analyte performance. Results: Different sigma scores were obtained for analytes on different total allowable error standard, only triglyceride showed world class performance along all the three total allowable error standards. However, creatine kinase &amp; total bilirubin showed world class on clinical laboratory improvement amendment (CLIA) and biological variable database (BVD). On the other hand, glucose, and aspertate amino transferase had poor or unacceptable performance on all the total allowable error standards. Conclusion: Six sigma is an excellent tool in evaluating the quality performance of clinical biochemistry laboratory analytical process, quality goal index can help identify the problem and westgard rules can help design individual quality improvement strategy.

Background: Clinical laboratories function to deliver accurate, reliable, and timely reported results which are used in decision-making, diagnosis, and monitoring. Sigma metrics help to assess analytic methodologies and provide benchmarks for laboratories to design protocols for internal quality control (IQC), address poor assay performance, and assess the efficiency of existing laboratory processes. Thus, this study was undertaken to estimate the coefficient of variation (CV%), bias%, and total allowable error (TEa) of quality control (QC) samples using EM 200, to compare the TEa using Clinical Laboratories Improvement Act (CLIA) guidelines, and to analyze the sigma metrics of level 1 QC samples. Aims and Objectives: The aims and objectives of the study are to estimate the CV%, Bias%, TEa of QC samples using EM 200, to compare the TEa using CLIA guidelines, and to analyze the sigma metrics of level 1 QC samples. Materials and Methods: A cross-sectional study was carried out in the Central Biochemistry Lab, Karwar Institute of Medical Sciences, Karwar. IQC data level 1 of 15 analytes was analyzed using EM200 for 3 months. CV% is calculated from internal quality data, whereas bias% is obtained from an external quality assurance program. Sigma metrics were calculated using bias% and CV%. TEa was calculated and compared with CLIA guidelines. Results: We have &lt; 3 sigma values (unstable, unacceptable) for urea, creatinine, BID, serum glutamic oxaloacetic transaminase, serum glutamate pyruvate transaminase (SGPT), protein, cholesterol, calcium, 3–6 (ideal) for glucose, uric acid, total bilirubin, alkaline phosphatase, albumin, high-density lipoprotein, and &gt;6 (excellent) for triglycerides. TEa observed less than or close to CLIA suggests quality requirement met, and TEa observed more than CLIA (urea, creatinine, BID, SGPT, protein, and calcium) suggests methodologies need evaluation. Conclusion: Sigma metrics help to assess analytical methodologies and augment laboratory performance. Each and every laboratory can use sigma metrics as a guideline for QC strategy and a self-assessment tool for proper functioning.

IntroductionThe aim of this work is to design a selection algorithm for total allowable error (TEa) source using a graphic tool that, by integrating internal (IQC) and external (EQC) quality control performances, enables the laboratory to evaluate which TEa source better fits the test analytical performance.Materials and methodsTwo analytical performance indicators (bias and Sigma metrics) were estimated for 23 biochemistry tests during 2016. Bias was estimated on the EQC, and Sigma metrics was calculated through the results obtained in the IQC. The Sigma metrics was charted as a function of the bias (TEa%). Following the proposed algorithm (considering the hierarchy in the Milan 2014 consensus), the TEa was evaluated depending on two areas. One area in the chart was defined as the objective area in which the used TEa is the appropriate one for the analytical performance of the test under evaluation. For any test located outside this area, a performance re-evaluation was required using another source of TEa.ResultsIn 19 out of 23 evaluated tests, the resulting analytical performance allowed for the selection of biologic variability as TEa source. In the four remaining cases, TEa sources of lesser hierarchy were selected.ConclusionThe graphic tool designed together with the proposed algorithm enabled the laboratory to standardize the selection procedure of the most appropriate TEa for the test analytical performance.

Sigma metrics analysis of serology screening assays to enhance quality and efficiency in New Zealand blood services

IntroductionGlycosylated hemoglobin (HbA1c) concentrations measured in clinical chemistry laboratories show large differences between their interlaboratory reported values. Laboratory measurements of quality performance should be based on quantitative data. The sigma metrics model provides an objective method for the assessment of current HbA1c assays and is useful in quality management planning. The aim of our study was to evaluate the analytical performance of the MQ-2000 PT HbA1c analyzer test results in the context of our operating conditions on the sigma scale.Materials and methodsThe coefficient of variation was determined from the calculated mean and standard deviation evaluated from internal quality control (QC) (N = 168 days) (Shanghai Huachen Biological Reagent Co. Ltd, China) data, and records of external quality data (KBUDEK, İstanbul, Turkey) measured in the period from May to November 2013 were used to determine the bias. The resulting data and total allowable error rate (TEA = 10%) from the Clinical Laboratory Improvement Amendments of 1988 (CLIA’88) were used to calculate the sigma level.ResultsThe calculated coefficient of variations (CVs) at the two levels, normal (QC1 = 36.6 ± 2.38 mmol/mol) and pathological (QC2 = 84.7 ± 2.68 mmol/mol), were 6.5% and 3.1%, respectively. The average bias between the external QC and MQ-2000 PT during the study period was 4.3%. The calculated average sigma value was 1.19.ConclusionsThe MQ-2000 PT HbA1c is a new analyser in the market; there is need for improvement and the method should be controlled with greater attention to ensure quality.

Sigma metrics, implemented by clinical laboratories, utilize the performance of assays to evaluate both precision and accuracy by considering the recommended total allowable error (TEa). The study aims to evaluate the performance of 48 analytes (chemistry and immunological) performed in a tertiary care hospital laboratory and compare them with the TEa recommended by CLIA 1988 and 2024 guidelines. This study, conducted at the Biochemistry Department of Medanta Hospital, Lucknow, Uttar Pradesh, India, utilizes retrospective data of the period January 2023 to December 2023 to analyze the performance of 48 assays in terms of precision (percentage of coefficient of variation (CV%) from the internal quality control (IQC) data) and accuracy (Bias %) from the monthly data of external quality assessment services (EQAS). 28 clinical chemistry and 20 immunoassays were performed on an automated VITROS XT 7600 Integrated chemistry analyzer. Sigma scores were calculated and analyzed using a standard formula, which includes Bias%, CV%, and TEa for each assay. Based on TEa source from CLIA 1988 and 2024, for clinical chemistry assays - 64%, 46% showed sigma performance score of &amp;#62;6.0; 27%, 43% showed sigma performance score between 3.0 to 5.9; and 9%, 11% of showed sigma performance score of below 3.0 respectively. For immunoassays, based on TEa source from CLIA 2024, 55% showed a sigma performance score of &amp;#62;6.0, 35% showed a sigma performance score between 3.0 to 5.9, and 10% showed a sigma performance score of below 3.0. This study showed that about 90% of clinical chemistry and immunoassays produced excellent results on the sigma scale. The analysis helped to identify the root causes of the low performance of a few assays with a sigma score below 3.0 and the performance improvement steps undertaken.

A comparative analysis of Sigma metrics using conventional and alternative formulas

Background Chemistry panels play a crucial role in determining a person’s general health and well-being. These tests evaluate electrolyte balance and the status of several major organs. The Comprehensive Metabolic Panel (CMP) and Lipid Panel constitute approximately 87% of the blood tests prescribed in medical laboratories. Laboratories monitor assay performance by performing proficiency testing using quality controls (QC). College of American Pathologists (CAP) data is useful for laboratories to compare their proficiency testing performance to other laboratories and instrument systems. Sigma metrics serve as a valuable tool for laboratories to assess the quality and accuracy of their tests. Sigma scores vary from 0 to 6, with the 6 score designated as world-class performance, 5 as excellent, 4 as good, 3 as marginal, and 2 as poor. The analytical performance of 16 clinical assays from Abbott, Beckman, Roche, and Siemens instrument systems were compared using Sigma metrics. Methods Sigma metrics of 16 chemistry analytes were compared across 4 instruments system using CAP data from three reports (C-A 2023, C-C 2023 and C-B 2022). The 4 systems evaluated were Abbott Alinity ci, Beckman AU, Roche COBAS c500/c303 and Siemens Atellica CH. Method imprecision was estimated for each assay using the cumulative coefficient of variation or standard deviation for each assay at a single QC concentration and calculating the average sigma. Since sigma levels can be concentration-dependent, an equivalent CAP sample concentration was chosen for each instrument system evaluation. Method bias was estimated by evaluating each method QC mean to the peer group mean. Sigma values were calculated for each method as Sigma = (TEa - |Bias|) / Precision using the allowable total error from 2014 Clinical Laboratory Improvement Amendments (CLIA) proposed acceptance limits. Average Sigma values were generated for each analyte and graded as an optimal Sigma score. Results The sigma scores varied across assays and instrument systems. Abbott demonstrated the best overall performance. Specifically, the Abbott Alinity system had the highest number (7/16, 44%) of assays at 5 sigma or higher, with Siemens showing the lowest number of assays at 5 Sigma or higher (1/16, 6%). Of the 16 assays evaluated, Abbott has the fewest assays at or below the 3 sigma threshold (5/16) when compared to Roche (7/16), Beckman (7/16), and Siemens (9/16). Sigma scores were comparable between the Abbott ARCHITCT and Alinity clinical chemistry systems. Conclusions Sigma metrics are useful in assessing the performance of clinical assays. Assays with high Sigma have better performance due to lower imprecision and higher accuracy. None of the instrument systems demonstrated optimal six sigma performance for all analytes. However, Abbott demonstrated the best overall performance.

Background(a) To evaluate the clinical performance of endocrine analytes using the sigma metrics (σ) model. (b) To redesign quality control strategies for performance improvement.MethodsThe sigma values of the analytes were initially evaluated based on the allowable total error (TEa), bias, and coefficient of variation (CV) at QC materials level 1 and 2 in March 2018. And then, the normalized QC performance decision charts, personalized QC rules, quality goal index (QGI) analysis, and root causes analysis (RCA) were performed based on the sigma values of the analytes. Finally, the sigma values were re‐evaluated in September 2018 after a series of targeted corrective actions.ResultsBased on the initial sigma values, two analytes (FT3 and TSH) with σ > 6, only needed one QC rule (13S) with N2 and R500 for QC management. On the other hand, seven analytes (FT4, TT4, CROT, E2, PRL, TESTO, and INS) with σ < 4 at one QC material level or both needed multiple rules (13S/22S/R4S/41S/10X) with N6 and R10‐500 depending on different sigma values for QC management. Subsequently, detailed and comprehensive RCA and timely corrective actions were performed on all the analytes base on the QGI analysis. Compared with the initial sigma values, the re‐evaluated sigma metrics of all the analytes increased significantly.ConclusionsIt was demonstrated that the combination of sigma metrics, QGI analysis, and RCA provided a useful evaluation system for the analytical performance of endocrine analytes.

For methods that report quantitatively, an assessment of their imprecision and bias should be assessed in the laboratory before their deployment into routine service. This study assessed these parameters of the HbA1C method on the STANDARD F2400® point of care analyzer. These parameters were further combined to generate sigma metrics for the method. An external quality assurance (EQA) material from the Randox International Quality Assessment Scheme (RIQAS) was analysed according to the EP15 protocol of the Clinical Laboratory Standards Institute in SYNLAB Nigeria Laboratory Quality Assurance Department. Estimates of precision and an assessment of bias were determined from the data which consisted of 5 replicates per day for 5 consecutive days. Precision estimates were compared with manufacturer-provided information and estimates of bias were compared with the verification interval for the target value provided by RIQAS. Sigma metrics were determined for total allowable error (TAE) of 8% and 10%. The grand mean (standard deviation) for the study was 4.95 (0.15) %. The Within-run CV and within laboratory CV were 1.28% and 1.86%, respectively. These were within the manufacturer claims of 1.70% and 1.90%, also respectively. The target value by RIQAS was 5.04(0.24) % with a calculated verification interval of 4.95 - 5.13%. The sigma metrics for the method at TAE of 8% and 10% were 4.8/3.3 and 6.3/4.3 within the run/within laboratory estimates respectively. The HbA1C method on the STANDARD F2400® Analyzer displayed performance characteristics that are consistent with manufacturer specifications and are above industry standard quality for a point-of-care device for HbA1C. These suggest that may be used to support routine monitoring of persons with diabetes mellitus in Nigeria.

Objectives: This study aimed to evaluate the analytical performance of coagulation tests (prothrombin time (PT/INR), activated partial thromboplastin time (aPTT), fibrinogen, and D-dimer) using the Six Sigma methodology, focusing on identifying areas for improvement to enhance healthcare quality. Methods: Internal quality control (IQC) and external quality control (EQC) data from September 2023 to February 2024 were collected for coagulation parameters analyzed by four Cobas T711 analyzers. Sigma values were calculated using IQC, EQC, and total allowable error (TEa) data. The sigma value for each parameter was calculated according to the formula "(TEa%-bias %)/CV%. " The outpatient sample analyzers were labeled A1, A2, and A3. The one that analyzes samples from emergency and intensive care patients was labeled B. Results: Across analyzers A1, A2, A3, and B, sigma values varied for different coagulation parameters. Notably, the D-dimer parameter consistently exhibited excellent performance (sigma >6) for all analyzers. In contrast, some analyzers showed poor performance for aPTT and PT parameters at level 1 (A1 and A3 for aPTT, B for INR). Fibrinogen performance varied, with some analyzers showing excellent performance (sigma >6) and others falling below acceptable levels. Conclusion: By identifying areas of low performance, particularly in aPTT and INR parameters, this study highlights the importance of continuous quality improvement in laboratory testing. Addressing issues identified through the Six Sigma methodology can enhance the reliability of laboratory results and ultimately improve patient care. Further research and initiatives focused on analytical process improvement are needed to achieve higher quality standards in laboratory testing.

Background We evaluated the analytical performance of four newly developed clinical chemistry assays [Creatinine, Uric Acid, Lactate, Lactate Dehydrogenase (LDH)] on the Abbott Alinity c and ARCHITECT c8000 system. These clinical chemistry assays incorporate a multianalyte calibrator, which has extended stability. For the enzyme assay LDH, both the multianalyte calibrator and a factor calibration are offered. Precision, linearity, method comparison and accuracy were assessed per CLSI guidelines. Sigma metrics were determined to examine the assay variability in relation to allowable performance limits. Methods Precision was assessed by measuring two levels of quality control material (Biorad Chemistry UA) and three pooled patient samples in replicates of five, twice per day, for five days. Method comparison was performed by measuring 134 serum/plasma specimens in duplicate on the new and current Alinity c or ARCHITECT methods. Linearity was evaluated using five to six levels of commercially available linearity material, measured in triplicate. Acceptability criteria were based on the Accreditation Canada Diagnostics (ACD) recommendations. Accuracy was determined in replicates of ten using external NIST standards. For assays where an external standard was not available, calibrators were used. The imprecision and bias from the target value were used to calculate the Sigma value. Total allowable error limits were obtained from the Clinical Laboratory Improvement Amendments (CLIA) and ACD guidelines. Results The results from the precision, accuracy, linearity, and method comparison studies are summarized in the table below. All the next generation chemistry assays demonstrated = 6 Sigma performance on the ARCHITECT and Alinity c systems based on CLIA guidelines. The performance of the factor calibration for the LDH assay was equivalent (data not shown). Conclusion The four next generation clinical chemistry assays (Creatinine2, Uric Acid2, Lactate2, Lactate Dehydrogenase2) demonstrated acceptable performance for precision, accuracy, linearity, and Sigma values on the Alinity c and ARCHITECT system. There was good agreement with the on-market Alinity c and ARCHITECT clinical chemistry assays.