Tools, Technologies and Training for Healthcare Laboratories

Patient-Based, Risk-Driven, Real-Time. The Adjective-Hyping of Quality Control

There's a race to develop new adjectives for Quality Control. Statistical Quality Control is out. So is Process-Focused. Patient-Focused Real-Time, and Risk-Based are the latest terms in vogue. But what does any of this actually mean to the laboratory?

Process-Focused vs Patient-Focused vs Risk-Based Quality Control

James O. Westgard, PhD
December 2022

What’s in a name? For example, “Quality Control”? It sounds great, because it implies that the practice of analyzing control materials controls quality. But, what is the level of quality is being achieved? Has it been specified in the design and planning of the testing process? Has it been measured in practice? If not, what is thea proper understanding of what we call QC?

Process-Focused vs Patient-Focused QC

Such a practice might better be understood as statistical process control because it focuses on process variation, with the objective of identifying and eliminating special causes of variation to achieve stable and predictable variation. As such, the actual level of quality is not known and the practice might be better described as arbitrary control. That doesn’t sound as good, does it?, and wWe would never want our customers and consumers to think that we don’t actually know the quality we control.

Quality Ccontrol suggests that the level of quality has been defined and designed into the process. In medical laboratories, that level of quality is often referred to as “intended use” for patient care. Achieving a state of stable statistical control of process variation is a prerequisite for statistical quality control, therefore it may be more correct to talk about the difference between process control and quality control. However, that is not the common usage in medical laboratories or even in other industries.

An important issue is how to incorporate the quality required for the intended medical application into the design of SQC strategies. Recall the SQC guidance from ISO 15189 [1]: “The laboratory shall design quality control procedures that verify the attainment of the intended quality of results.” That requires the laboratory to first define the quality required for the intended medical use of a test and second to apply that requirement in the selection/design of appropriate control rules and number of control measurements. Process characteristics that must be considered include the precision and bias observed for the measurement procedure and the probabilities for rejection expected for the control procedure.

Instrument-focused vs Patient-focused QC

A recent discussion described the current laboratory QC practice as focused on measurement technology and detection of instrument errors rather than being focused on patient outcomes and detection of medically important errors [2]. The implication is that laboratories should be designing QC to reduce the risk of reporting erroneous patient test results. The distinction, according to the authors, is that laboratories should define the requirement for intended use in terms of an allowable Total Error (TEa) and then select the control rules and numbers of control measurements to detect medically important errors.

Guidelines for designing SQC to detect medically important errors were first introduced in the 1990s [3] and a QC planning process was recommended in 2006 in the 3rd edition of the CLSI C24 guidance for statistical QC practices [4]. Based on a defined TEa goal, the imprecision and bias observed for a method, and the known rejection characteristics (probabilities for error detection, Ped, and false rejection, Pfr) for different QC procedures, optimum control rules and an appropriate number of control measurements can be selected for a QC event. Power function graphs were included in C24-A3 to provide a practical tool for selecting appropriate control rules and numbers of control measurements.

Continuous Production Processes vs Batch Processes

What is different in today’s state of practice is the need to adapt this approach for the continuous production instruments that are the workhorses in highly automated laboratories. These instruments produce test results continuously and the concern is where and when to analyze control samples. The CLIA regulatory requirement is only that two levels of controls be analyzed per day, which may mean every 8 to 24 hours. Typically this leads to a practice of analyzing controls at the beginning and end of an analytical run. The initial controls should assure that correct test results are being produced at the beginning of the run and the final controls should ensure quality test results have continued to be produced throughout the analytical run. The CLSI C24-Ed4 document furthermore recommends “bracketed operation” whereby the test results between consecutive control events are reported when both QC events are found to be “in-control.”

With continuous production processes, long runs can be a problem when patient test results are reported continuously (unbracketed operation) and there is an out-of-control QC event after patient test results have already been reported. In principle, the control problem must be investigated, and corrective action taken to bring the testing process back into conformance. Then patient samples may have to be re-analyzed to assure test results are correct. Long analytical runs may lead to correcting patient reports long after they were first reported.

The reality of the continuous production approach of modern analyzers is that results go out with an assumption of correctness, rather than a verified (bracketed) correctness. The naked risk goes out to patients, and the laboratory crosses its fingers.

In this situation, the frequency of QC events should be optimized to minimize the risk of reporting erroneous patient results by utilizing Parvin’s patient risk model [5]. This type of QC may more properly be referred to as a risk-based SQC strategy. As such, it specifies the control rules, number of control measurements, and the frequency of QC events (specified in terms of run size, or how many patient samples analyzed between QC events). A “roadmap” for planning such risk-based SQC strategies is provided in the 4th edition of CLSI C24 document [6].

Unfortunately, C24-Ed4 does not provide sufficient details to implement a practical planning process for risk-based SQC strategies. The mathematical/statistical calculations required for the risk model are complicated and likely limit the usefulness to the laboratory. But alternative graphical tools have been described in the literature [7-10] and simple on-line calculators are also available on this website [11]. Thus, even small laboratories can implement “state of the art” risk-based QC procedures, as documented in the recent literature [12].

What’s the point?

What is the state of practice in US laboratories? If we consider 1st generation QC to be process control, 2nd generation to be patient-focused quality control, 3rd generation to be risk-based SQC strategies, and 4th generation to be van Rossum’s Integrated Quality Assurance and Control (IQAC) that also includes patient-data QC procedures [13], where would you rank your laboratory? Based on a 2016 survey of academic laboratories that should represent best practices [14], most US labs do not define the quality required for intended use or do not use such a requirement to select or design SQC, thus they mainly practice statistical process control. They’re barely in the 1st generation of QC practices. Furthermore, they primarily employ a one size fits all tests on their high-volume multi-test systems, even though that practice is known to be suboptimal. And to further complicate applications, it is common that this “one size QC” employs a Repeat:2SD control rule, in spite of the expected high number of false rejections and the logistical issues of repeating controls [16]. Put that all together, some laboratories are stuck in the primordial ooze of QC practices, nowhere near the 21st century possibilities available to them.

This is disappointing, particularly when the CLSI C24 document provides the leading consensus on good laboratory SQC practices. It may be that the minimal requirements of the US CLIA regulations have drowned out over-counter-balanced those the recommended good best practices, as well as related standards for accreditation (from organizations “deemed” to provide standards equivalent to CLIA). It is also likely that the CLSI C24-Ed4 document needs to be improved and made practical – most helpfully by making it readable - and readable for laboratory supervisors and bench level staff [9].


  1.  ISO 15189:2012. Medical laboratories – particular requirements for quality and competence. International Organization for Standardization (ISO), Geneva.
  2. Yundt-Pacheco J, Parvin C. The focus of quality control strategies should be on patient outcomes, not technology. MLO 2022;54(No.12):32-34.
  3. Westgard JO. A QC planning process for selecting and validating statistical QC procedures. Clin Biochem Revs 1994;15:156-164.
  4. CLSI C24-A3. Statistical Quality Control for Quantitative Measurement Procedures: Principles and Definitions – Third Edition. Clinical and Laboratory Standards Institute, 950 West Valley Road, Suite 2500, Wayne PA, 2006.
  5. 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.
  6. CLSI C24-Ed4. Statistical Quality Control for Quantitative Measurement Procedures: Principles and Definitions. 4th ed. Clinical and Laboratory Standards Institute, 9950 West Valley Road, Suite 2500, Wayne PA, 2016.
  7. Yago M, Alcover S. Selecting statistical procedures for quality control planning based on risk management. Clin Chem 2016;62:959-965.
  8. Bayat H. Selecting multi-rule quality control procedures based on patient risk. Clin Chem Lab Med 2017;55:1702-1708.
  9. Bayat H, Westgard SA, Westgard JO. Planning risk-based SQC strategies: Practical tools to support the new CLSI C24Ed4 guidance. J Appl Lab Med 2017;2:211-221.
  10. Westgard JO, Bayat H, Westgard SA. Planning risk-based SQC schedules for bracketed operation of continuous production analyzers. Clin Chem 2018;64:289-296.
  11.  Online QC Frequency Calculators:
  12. Lukic V, Ignjatovic S. Integrating moving average control procedures into the risk-based quality control plan in small-volume medical laboratories. Biochem Med (Zagreb) 2022;32(2).
  13. Van Rossum HH. Technical quality assurance and quality control for medical laboratories: a review and proposal of a new concept to obtain integrated and validated QA/QC plans. Crit Rev Clin Lab Sci.
  14. Rosenbaum MW, Flood JG, Melanson SEF, Baumann NA, Marzinke MA, et al. Quality control practices for chemistry and immunochemistry in a cohort of 21 large academic medical centers. Am J Clin Pathol 2018;150:96-104.
  15. Westgard JO, Westgard SA. Establishing evidence-based Statistical Quality Control Practices. Am J Clin Pathol 2019;151:364-370.
  16. Westgard JO, Bayat H, Westgard S. Chapter 9. Evaluating Repeat:2s QC Practices. Advanced QC Strategies: Risk-Based Design for Medical Laboratories. Madison, WI:Westgard QC, Inc., 2022, pp 177-188.