A survey finds daily out-of-control quality control events reported by 33% of labs worldwide and 45% in the US, raising concerns about current QC practices amid rising operational pressures.
By Sten Westgard, MS, director of client services and technology, Westgard QC
The core competency of a clinical laboratory is providing reliable test results. Few dispute this. Consistent quality control (QC) is essential to continuously verify analytical accuracy.
But what is the quality of QC? Are current QC practices best suited to today’s high-volume, resource-strained, staffing-challenged laboratories?
Increasingly, the answer to the former question is “Poor.” And the answer to the latter question is “No.”
Out-of-control Events Are on the Rise
QC is supposed to indicate when an analytical process is no longer providing reliable results—an occurrence known as an out-of-control (OOC) event. These OOC events can occur due to several reasons, including instrument malfunctions, reagent deterioration, temperature fluctuations and technician error. However, they can even more easily occur due to false rejection created by the application of inappropriate QC rules and limits.
In 2025, Westgard QC conducted a survey of more than 1,200 laboratories across more than 110 countries, including more than 400 labs in the United States. This 30-question survey investigated the real-world implementation of QC across a diverse range of laboratories, including, public and private organizations, single-site and multi-network laboratories and disciplines spanning chemistry, hematology and more.
According to the Westgard survey, one in three (33.3%) laboratories worldwide now experience an OOC event every day, if not multiple times per day. In the US, this number is even higher: 45% of laboratories are out –of control every day, if not multiple times per day. If we narrow the field down to just the chemistry sections of US laboratories, 55% of those labs are out of control every day, if not multiple times per day.
These figures are as concerning as they are paradoxical. Laboratories have never had more advanced, capable instruments, and yet these same instruments seem to be out of control more than ever. The same labs that record constant alarms nevertheless continue to report patient results daily. How can this be? It only makes sense if the vast majority of these OOC events are not real errors but are instead false alarms.
Frequent OOC events point to a deficiency within QC practices. When a laboratory’s internal QC creates so many out-of-control alarms, the QC itself evolves to present a greater problem than the test errors it is attempting to detect. Ineffective QC processes hamper laboratory operations, introducing key inefficiencies, such as excessive troubleshooting, inconsistent documentation and delayed result reporting, all elevating total operational costs and reducing throughput.
In today’s operating environment, where time and talent are scarce, and financial constraints are intensifying, frequent OOC events stress laboratory resources beyond their breaking point.
Shortcomings of Today’s QC Practices
QC remains the bedrock of laboratory operations, yet the actual practice of QC seems to be lodged in the Stone Age. QC procedures that were developed nearly half a century ago are often still in active use in modern laboratories.
Specifically, control limits that are calculated as the mean plus and minus 2 standard deviations (SD), commonly referred to as the 1:2s rule, when applied with two or three control levels, remain the predominant approach in most laboratories worldwide, despite decades of evidence that they generate high false rejection rates. The 1:2s rule using 2 SD control limits generates about a 9% false rejection rate when two control measurements are evaluated per test run. With three control measurements, that false rejection rate rises to 14%.
In the United States, two-thirds (63.9%) of laboratories use the 1:2s rule or 2 SD control limits on all or some of their QC tests. The use of 2 SD limits extends to the majority of laboratories around the globe, for all or some of their tests as well. The use of 2 SD control limits is a suboptimal practice that gradually became commonplace over time. As the constant false outliers generated by the 1:2s rule metastasized across testing menus and multiple platforms, laboratories, accepted more outliers, until it became “normal” to expect at least one, if not more, QC outliers every day.
That is a lot of false alarms we have baked into our laboratory QC system. It is a level of noise (eg, measurement error) that we probably would not accept in any other operational process. Worse still, we have always known that constant alarms lead to diminished responsiveness.
Like The Boy Who Cried Wolf, when QC generates daily false alarms, overall trust in the QC process is diminished—an occurrence known as alarm fatigue.
Instead of fully troubleshooting the OOC event, weary labs resort to a tried-and-true and deeply flawed workaround: the infamous control repeat.
Repeated Controls: A Natural Consequence of Alarm Fatigue
The United States leads the world in repeating controls, with 88.8% of laboratories taking this step instead of fully troubleshooting.
But one do-over often leads to another (if one repeat is OK, why not two, or three, or…). As shown in Figure 1, about 75% of US laboratories will run a new control after a failed repeat, and about 23% will repeat the new control if that fails, too. Anecdotally, we have encountered laboratories that repeat and repeat and repeat and repeat until they finally get the “in” result they are seeking. (Simultaneously ignoring all of the “out” results that came before.)
Figure 1. USA: How do you trouble-shoot an out-of-control QC?i
In other words, laboratories start off with the wrong limits—remember, 2 SD is too tight— and then they respond to the excessive outliers with the wrong response: repeating the controls. But this is not a case where two wrongs make a right. Two wrongs, in this case, double the waste.
Indeed, the waste grows and grows as extra controls are run, extra calibrations are performed, more reagent is consumed, longer delays are created when patient results are held pending troubleshooting, and, worst of all, the laboratory’s most valuable resource—the time and effort of the laboratory scientist—is consumed by these unnecessary actions.
Perhaps the laboratory of the past was able to absorb these costs. The modern laboratory, with its ever-expanding menu, ever-increasing volume and ever-shrinking labor force with diminishing QC expertise, cannot afford to perpetuate this wasteful cycle.
Luckily, nothing about using 2 SD is inevitable. The 1:2s rule is a tradition, yes, but it is not mandatory. Other, better choices can be made for QC practices.
Enter the Westgard Rules
In 1981, the Westgard system introduced a multirule QC procedure that dramatically reduced the false rejection rate of QC. Instead of 9% to 14%, the “Westgard Rules”, as they came to be known, generated only 3% to 4% false rejection, while still offering robust error detection. These rules were not patented or commercially exclusive, but given freely to the laboratory world, and were consequently, rapidly adopted throughout the world. Most QC programs today have some form of Westgard Rules included by default.
Figure 2. Original Westgard Rules
Alas, the respite generated by the Westgard Rules was only temporary. While the 1:2s rule was implemented as a “warning rule” in the original Westgard Rules, slowly that warning rule mutated into a repeat rule. The Westgard Rules have never advocated repeating controls. However, laboratories seeking a QC shortcut started (or returned to) repeating controls for everything outside 2 SD control limits, rather than working through troubleshooting and the other rejection rules.
Westgard Rules and QC for the Modern Laboratory
In contrast to the backward drift of QC practices, the analytical performance of instruments has leaped ahead. Today’s instruments, particularly biochemistry platforms, demonstrate markedly improved precision, with better traceability and accuracy. Some methods are so robust, in fact, that the QC of the past is entirely inappropriate.
The natural question arises: How does a laboratory know how much QC is necessary for any given test? After all, if the chemistry method of 2026 is clearly better than one from 1981, why would the same set of Westgard Rules be needed to monitor it?
There is a seductive simplicity in applying a “one-size-fits-all” approach to QC rules and tests. But that will undoubtedly over-QC some methods and under-QC others. The laboratory needs a tool to help customize QC on a test-by-test basis.
Introducing the Analytical Six Sigma Metric
In 2001, the Westgard system adapted Six Sigma theory to statistical QC in the medical and clinical laboratory. The analytical Sigma metric was introduced, which allowed the laboratory to benchmark the performance of a test on the universally accepted Six Sigma scale.
Then, in 2015, Westgard introduced a simplified set of Westgard Rules that included the Six Sigma benchmarks, now known as “Westgard Sigma Rules”.
Figure 3. Westgard Sigma Rules
Understanding the analytical Sigma metric of a method not only tells the laboratory the reliability of the test; it also dictates the amount of QC that is necessary to monitor it. Higher Sigma? Less QC, with fewer rules and wider limits. Lower Sigma? More QC, with more rules and tighter limits. A Six Sigma method only needs to use the 1:3s control rule with two controls.
Knowing you have a Six Sigma method means you can reduce more than 90% of the false alarms that plague your traditional 1:2s rule. You still retain all the necessary error detection to identify important errors, but you are no longer burdened by the chronic noise created by too-tight limits.
Table. Six Sigma Savings (or The Table of Noise Reduction)
| Analytical Sigma Metric | Alarm reduction from 1:2s with 2 controls | Alarm reduction from 1:2s with 3 controls |
| 6 Sigma | 94.4% | 96.4% |
| 5 Sigma | 88.9% | 92.9% |
| 4 Sigma | 77.8% | 85.7% |
| 3 Sigma | 66.7% | 78.6% |
| <3 Sigma | 66.7% | 78.6% |
For example, in a recent analysis of peer group median performance for clinical laboratories, Westgard QC found that using Six Sigma metrics for Amylase, Total Bilirubin, HDL and Triglycerides, a majority of the major diagnostic platforms can reduce their QC rules, slashing hundreds of outliers a year from their troubleshooting.
Automated Calculations Speed Sigma Metric Savings
When analytical Sigma metrics were first introduced to medical laboratories over 25 years ago, all the work needed to be done manually, using hand-crafted spreadsheets or specialized software. Today’s major control vendors provide Sigma metric calculations in their software as a matter of course.
Laboratories no longer need to seek out Sigma calculations; these metrics are often already a feature of their peer group software. So, the obstacle now to implementing Sigma metrics and realizing the significant savings is not technical, it is psychological. The greatest barrier laboratories face now is their own fear of statistics, their own adherence to outdated traditions of QC and their false comfort in the status quo.
Figure 3. LabLink 360 View of Westgard Sigma Rules
Conclusion
The last 50 years have seen few advances in QC, leaving us with an antiquated set of rules and limits.
With an ever-expanding testing menu and volume, the laboratory can no longer tolerate the excessive waste created by false rejections and unnecessarily repeated controls.
For laboratories that embrace better QC rules, including analytical Six Sigma metrics, the future holds fewer repeats, fewer levels needed, fewer controls run, fewer recalibrations needed, fewer hours spent chasing ghosts, and fewer hairs being pulled out of frustrated heads. The benefits of this new approach save money on controls, calibrators, reagents, staff time, and turn-around time.
This is the only thing laboratories should be repeating today:
Stop repeating controls unnecessarily.
Stop repeating controls unnecessarily.
Reference
[i] Westgard QC, Inc. 2025 Great Global QC Survey. Conducted in partnership with Thermo Fisher Scientific. Westgard QC, Inc; 2025. Accessed August 10, 2026. https://www.westgard.com/qc-applications/basic-qc-practices/2025-global-qc-survey.html
