Quality Requirements and Standards

The IFCC has issued a new guideline on how to properly implement IQC. Surprisingly, it still supports the use of Westgard Rules and analytical Sigma-metrics. Not surprisingly, a growing emphasis (and confusion?) about the use of measurement uncertainty.

2025 IFCC Recommendations for Ensuring IQC Practices in Medical Laboratories

James O. Westgard, PhD
March 2025

Introduction - a personal history

I became interested in Statistical QC while on sabbatical at Uppsala University in Sweden in 1976-77. At that time, IFCC had just published its initial draft [1] of guidance documents on quality control in clinical chemistry. These documents discussed principles and terminology, assessment of analytical performance, calibration and control materials, internal quality control, and external quality control. Today, almost 50 years later, IFCC has published new guidelines to address QC practices and requirements identified in ISO 15189:2022 [2]. Few lab scientists today remember the original IFCC guidelines, but they were essential for understanding the state of laboratory quality management at that time.

The new recommendations can be found here:

Giannoli J_M, Vassault A, Carobene A, Liaudet AP, Blasutig IM, Dabla PK, Lin J, Thomas A, Tesser Poloni JA, Meng QH, Amann EP. IFCC Laboratory Medicine Task Force on Global Lab Quality (TF-GLQ). Ensuring internal quality control practices in medical laboratories: IFCC recommendations for practical applications based on ISO 15189:2022 Clin Chim Acta 2025 (doi: https://doi.org/10.1016/j.cca, 2025, 120240.

The TF-GLQ should be commended for their efforts and the accomplishment of agreement on a set of recommendations. The current trends in IQC are leaning toward outright abolishment: get rid of the Westgard Rules, get rid of Six Sigma, get rid of traditional IQC charts, get rid of traditional controls, replace it all with something else (universal PBRTQC, constant EQA, new uncertainty controls among the many proposed replacements). That the committee maintained a commitment to the fundamentals of quality is a testament to their dedication, or stubbornness.

In 1976, I also participated in a meeting on IQC that was held in a small hotel outside of Munich, Germany. I was attending on behalf of Professor Carl-Hendric deVerdier who had been invited to speak but had become seriously ill. I remember it well, in part because it was after an October Fest weekend and some of my fellow train travelers made heroic dashes from the train to the edge of the platform at stops along the way. I also remember that the hotel produced its own beer and wine, which were available free in the lobby in the evenings to stimulate discussions.

As a young clinical chemist, it was intimidating to find myself in the midst of these learned Clinical Chemists who had strong feelings about QC practices. Nonetheless, the meeting provided the first opportunity to discuss the work we had begun in Uppsala to assess the probabilities of rejection for different QC procedures [3]. That work was fundamental to the development of computer simulation tools for assessing performance characteristics in the form of power function graphs [4], recommendations for combined Shewhart/Cusum charts [5] and the multirule QC procedure that eventually became known as Westgard Rules [6].

Purpose of new IFCC IQC Recommendations

IFCC identifies the primacy of ISO 15189:2022 as the global guideline for quality management in medical laboratories. The evolution of the original IFCC guidance to ISO 15189 is not clear, at least to many of us outside of Europe. The original documents were drafted by clinical chemists from Germany (Buttner), England (Broughton), USA (Boutwell), Canada (Borth), and Australia (Bowyer). It is not known if it was a requirement or coincidence that the names of all committee members began with B, but it was certainly a statistical outlier.

ISO documents are often said to be written at a high level, which means the documents describe “what” needs to be achieved but not “how” to do it. Sometimes they refer to this as a "horizontal" standard, not a "vertical" standard. A vertical standard would go deep into the details, while the horizontal standard would simply identify priniciples and abstract goals, leaving out the details of how to achieve them (and in theory, giving laboratories the flexibility to find their own ways to accomplish goals). That is the reason for the new IFCC guidance on “ensuring” IQC practices. Implementation requires additional knowledge and skills.

In this discussion, I consider two important issues: how to plan IQC strategies and how to determine Measurement Uncertainty (MU).

IQC Planning

The original requirement for IQC in ISO 15189:2012 was a rather short statement:

5.6.1 The laboratory shall design internal quality control systems that verify the attainment of the intended quality of results. It is important that the control system provide staff members with clear and easily understood information on which to base technical and medical decisions. Special attention should be paid to the elimination of mistakes in the process of handling samples, requests, examinations, reports, etc.

The latest 2022 version (ISO 15189:2022(E)) expands that description in section 7.3.7.2 on Internal Quality Control:

a) The laboratory shall have an IQC procedure for monitoring the ongoing validity of examination results, according to specified criteria, that verifies the attainment of the intended quality and ensures validity pertinent to clinical decision making.

1) The intended clinical application of the examination should be considered, as the performance specifications for the same measurand can differ in different clinical settings.
2) The procedure should also allow for detection of either lot-to-lot reagent or calibrator variation, or both, of the examination method. To enable this, the laboratory procedure should avoid lot change in IQC material on the same day/run as either lot-to-lot reagent or calibrator change, or both.
3) The uses of third-party IQC material should be considered, either as an alternative to, or in addition to, control material supplied by the reagent or instrument manufacturer.

While the additional description is welcome, there may still be a need for further details. The earlier guidance stated the laboratory “shall design internal quality control systems that verify the intended quality of results” whereas the more recent guidance says the laboratory “shall have” such an IQC procedure. The earlier guidance made it clear that the laboratory should be involved in the design of its own QC procedures. The new guidance suggests that there is an existing supply of QC procedures that the laboratory already possesses, or perhaps is supplied to them. 

Here’s where the new IFCC guidance document provides an answer with its recommendation that “laboratories must establish a structured approach for planning IQC procedures, including the number of tests in a series and the frequency of IQC assessments.”

The laboratory must determine both the frequency of IQCs and the size of the series, which refers to the number of patient sample analyses performed for an analyte between two IQC events. While the Sigma level serves as a valuable tool for assessing the robustness of the method, additional factors must be considered as part of a comprehensive risk analysis:

• The clinical significance and criticality of the analyte.
• The time frame required for the result release and subsequent use.
• The feasibility of re-analyzing samples, particularly for tests with strict pre-analytical requirements (e.g. blood gas analysis, where re-testing may not be possible.)

The discussion of QC frequency reflects the latest scientific discussions.

There are several pages that discuss critical aspects of the IQC planning process:

• Definition of IQC frequency
• Evaluation of the robustness of the method [using Sigma-Metrics]
• Scheduling frequency/establishing the series and critical events
• Defining acceptability criteria
• Establishing target values
• Establishing Levey-Jennings charts
• Statistical control rules

These factors interact to provide appropriate IQC procedures or systems to verify the attainment of intended quality and provide assurance that test results are correct for clinical use. Fortunately, there are many detailed examples in the scientific literature beginning with Parvin’s patient risk model [8] which is the key for applying risk analysis to determine the IQC frequency and the appropriate run size. Also, there are graphical and online calculator tools that make it easy to apply Parvin’s model [9-14] and provide quantitative answers for the frequency of IQC and the appropriate run size (number of patient samples between control events). The new IFCC guidance is a critical step in the right direction.

Evaluation of Measurement Uncertainty (MU)

How to determine MU has been the subject of many papers for at least the last quarter-century and it remains an issue today. While there now seems to be agreement on a “top-down” approach that uses IQC and EQA data rather than a “bottom-up” approach that estimates the uncertainty of each variable or factor involved in the measurement process, there is still a major issue related to how bias should be handled. In general, metrologists argue that bias should ideally be eliminated, corrected if possible, and disregarded when necessary. The issue of bias is actually at the core of the arguments about using the Total Analytical Error model vs the Measurement Uncertainty model for quality management in medical laboratories.

In the 2012 version of ISO 15189, the requirement was stated in section 5.6.2, as follows:

• The laboratory shall determine the uncertainty of results, where relevant and possible. Uncertainty components which are of importance shall be taken into account. Sources that contribute to uncertainty may include sampling, sample preparation, sample portion selection, calibrators, reference materials, input quantities, equipment used, environmental conditions, condition of the sample and changes of operators.

There was no additional guidance for what was an impossible task. In truth, “where relevant and possible” gave the laboratory an escape hatch. As long as you claimed mu was not relevant and impossible, you could skip it.

In the 2022 edition of ISO 15189, the requirement is found in section 7.3.4 which states (in part):

• The MU of measured quantity values shall be evaluated and maintained for its intended use, where relevant.
• The MU shall be compared against performance specifications and documented.
• MU evaluations shall be regularly reviewed. 
• MU information shall be made available to laboratory users on request.
• MU should be taken into consideration when performing verification or validation of a method, when relevant.

MU is to be determined by the laboratory, compared to performance goals or specifications for allowable MU, considered in verification and validation of methods, and be made available to users upon request. In addition to how to determine MU, there are serious issues related to its use and interpretation. The escape hatch is now "where relevant." If you can plausibly explain why mu is not relevant, you still don't have to do it. But take that only as a temporary reprieve, there are serious pressures coming to make it completely mandatory, even when it isn't relevant or possible.

In considering the new IFCC guidance, first the guidance is brief, consisting of approximately 1 page of text in section 15 on the interpretation and estimation of measurement uncertainties, which includes the following:

• This section guides the calculation and interpretation of MU, including the identification of contributing factors such as imprecision and bias.

• Care should be taken not to confuse total error with MU. TE_A is calculated from the intra-subject (CV_I) and inter-subjects (CV_G) BV [biological variation] estimates, while the MU is an analytical uncertainty based on the quadratic combination of two terms: impression and bias. Bias should, in principle, be eliminated and all the remaining sources of variation added linearly as variances.

• A measurement result therefore can comprise two uncertainties: the uncertainty associated with bias correction and the uncertainty due to imprecision. For the bias, one can use the mean bias (with the standard deviation of the bias) or the maximum bias. Nevertheless, the major component of uncertainty is imprecision. So where the bias is deemed acceptable from data from EQA, or so small as compared to precision CV, it could be permissible to calculate the MU considering only two times the standard deviation (or CV) calculated within a long period. Results of IQC can be used for the evaluation of precision.

• The MU shall be evaluated according to the clinical interpretation required by clinicians, considering clinical needs.

So here are the possibilities for a “top-down” calculation of MU (where MU here represents a 95% estimate of expanded uncertainty) using data from IQC and EQA:

• In principle, MU is composed of the imprecision and bias, added together as variances, i.e., MU = 2√(CV² + Bias²)
• When Bias is small and not clinically significant, MU = 2*CV
• When Bias is significant but corrected, MU = 2√(CV² + UBias²), where UBias is the uncertainty related to the estimation of Bias (from an EQA survey)

This estimation of MU is still incomplete because the uncertainty of the reference and calibration (traceability information that must be supplied by the manufacturer as u_ref and u_cal) has not yet been included [15]. Even more interesting, the most up-to-date recommendation is to analyze a separate control for at least 60 days to estimate MU [16] is not incorporated in the IFCC guidance. There is a continuing lack of consensus about how to estimate and apply MU in medical laboratories. As we have seen with TE_A goals, the greater the number of recipes, the greater the confusion in the laboratory.

What’s the point?

The new IFCC recommendations provide much needed guidance, particularly the need to implement an IQC planning process. There are still serious issues in the daily practice of IQC, particularly in the design and planning of SQC strategies.  However, the recommendations demonstrate a continuing lack of consensus around the determination of Measurement Uncertainty and what to do with the results. While estimation of MU is required by ISO 15189:2022 (where relevant) for accreditation of laboratories, further guidance is needed to make MU useful and practical in medical laboratories. If mu continues to lag in relevance, there will be continuing lack of useful implementation in laboratories.

References

1. Buttner J, Borth R, Boutwell JH, Broughton PMG. International Federation of Clinical Chemistry provisional recommendation on quality control. Clin Cim Acta 63:F25 1976.
2. Giannoli J_M, Vassault A, Carobene A, Liaudet AP, Blasutig IM, Dabla PK, Lin J, Thomas A, Tesser Poloni JA, Meng QH, Amann EP. IFCC Laboratory Medicine Task Force on Global Lab Quality (TF-GLQ). Ensuring internal quality control practices in medical laboratories: IFCC recommendations for practical applications based on ISO 15189:2022 Clin Chim Acta 2025 (doi: https://doi.org/10.1016/j.cca, 2025, 120240.
3. Westgard JO, Groth T, Aronsson T, Falk H, deVerdier C H. Performance characteristics of rules for internal quality control: Probabilities for false rejection and error detection. Clin Chem 1977;23:1857 67.
4. Westgard JO, Groth T. Power functions for statistical control rules. Clin Chem 1979;25:863 69.
5. Westgard JO, Groth T, Aronsson T, deVerdier C H. Combined Shewhart cusum control chart for improved quality control in clinical chemistry. Clin Chem 1977;23:1881 87.
6. Westgard JO, Barry PL, Hunt MR, Groth T. A multi rule Shewhart chart for quality control in clinical chemistry. Clin Chem 1981;27:493 501.
7. ISO 15189:2022(E). Medical laboratories – Requirements for quality and competence. ISO, Geneva.
8. Parvin CA. Assessing the impact of the frequency of quality control testing on the quality of reported patient results. Clin Chem 2008;54:2049-2054.
9. Yago M, Alcover S. Selecting statistical procedures for quality control planning based on risk management. Clin Chem 2016;62:959-965.
10. Bayat H. Selecting multi-rule quality control procedures based on patient risk. Clin Chem Lab Med 2017;55:1702-1708.
11. Bayat H, Westgard SA, Westgard JO. Planning risk-based SQC strategies: Practical tools to support the new CLSI C24-Ed4 guidance. J Appl Lab Med 2017;2:211-221.
12. Westgard JO, Bayat H, Westgard SA. Planning risk-based SQC schedules for bracketed operation of continuous production analyzers. Clin Chem 2018;64:289-296.
13. Westgard JO, Bayat H, Westgard SA. Planning SQC strategies and adapting QC frequency for patient risk. Clin Chim Acta 2021;523:1-5.
14. Westgard SA, Bayat H, Westgard JO. A multi-test planning model for risk based statistical quality control strategies. Clin Chim Act 2021;523:216-233.
15. Westgard JO, Bayat H, Westgard SA. Determining MU from QC data. Chapter 14 in Advanced QC Strategies: Risk-Based Design for Medical Laboratories. Madison WI:Westgard QC, Inc. 2022.
16. Braga F, Pasqualetti S, Aloisio E. Panteghini M. The internal quality control in the traceability era. Clin Chem Lab Med 2020;59:291-300.