Category: Practice & Regulations | Estimated reading time: 7 minutes
Questions That Must Be Faced Honestly
After building a technical QC system — from preliminary testing to Sigma Metric — there is one question that must be faced honestly:
Is everything we have built enough?
The answer is surprising, and a bit uncomfortable: not always.
A Scenario That Looks Perfect — but Isn't
Imagine a laboratory that has never had a QC rejection for 3 consecutive months. Values always fall within the ±2SD range. But suddenly, there is a complaint from a doctor: results for the same patient differ significantly between shifts. After investigation, it turns out that two medical laboratory technologists (MLTs) are using different techniques — and no one has ever checked it.
This scenario is not about a failure of the QC system. The QC is running correctly — control materials are measured, Levey-Jennings charts are plotted, Westgard rules are applied, everything is within range. Nothing is technically wrong.
What failed is something different: consistency of procedures between operators — and this is something that daily QC is not designed to detect. This is not a weakness of QC, but rather the natural limitation of what QC can do. Understanding that limit is the first step toward building a truly comprehensive system.
What Daily QC Fails to Catch
There are several important things that fall outside the radar of daily QC, and these three deserve serious reflection by every laboratory.
QC only checks whether the control material is within range — not consistency between analysts. Control materials are homogenous samples, prepared in the same way, and often measured by the same operator or with highly standardized protocols. They do not represent the variation that occurs when two different MLTs process patient samples with different techniques — variations in pipetting, homogenization, or reading incubation times. This inter-operator variation is a very real source of error in the laboratory, but it is almost never detected by standard control material-based QC systems.
Operator variation, pre-analytical conditions, and procedural consistency are also not captured. Daily QC lives entirely in the analytical phase. It does not see what happens before the sample enters the instrument — patient identification, tube conditions, time between collection and processing, or transportation temperature. These are all sources of pre-analytical error that, if they occur, will not be detected by any QC system.
A study by Plebani & Carraro (1997) found that 68% of laboratory errors occur in the pre-analytical phase — a phase entirely beyond the reach of analytical QC systems.
If maximizing the capabilities of QC as a detection system has already been done, the next step is to build layers outside of QC that prevent problems before they even need to be detected.
The QC Paradox That Must Be Understood
There is an interesting paradox in the world of laboratory QC that is rarely discussed explicitly:
The laboratory that most frequently rejects QC is not necessarily the most problematic laboratory.
A laboratory with highly sensitive QC — using tight limits, many Westgard rules, and high frequency — will produce more rejections. But that is precisely what makes it safer: problems are detected earlier and more often.
Conversely, a laboratory whose QC is always "green" can be in two very different conditions:
- The measurement system is indeed very stable and of high quality, or
- The QC system is too loose — control limits are too wide, or the rules applied are not sensitive enough.
QC that is always green without a single rejection over a long period should actually be viewed with suspicion — not celebrated.
Westgard (2016) reminds us that the false rejection rate and the error detection rate are two sides of the same coin — a QC system that minimizes false rejection without considering the error detection rate is taking the easy way out at the expense of patient safety.
Stepping Outside of Daily QC
There are two major areas that complement what the daily QC system has built:
Advanced Quality Instruments — tools that work outside or above standard QC systems: detecting more subtle trends, positioning our performance in a broader context (e.g., through inter-laboratory comparisons/EQA), and identifying problems not caught by Westgard rules.
Quality Assurance (QA) — a system that works before QC: building optimal conditions where measurements occur, so that QC rarely needs to detect problems because they were prevented from the start — covering procedure standardization between operators, pre-analytical phase control, and continuous training.
Closing
QC that is always green is good news — but not an absolute guarantee. It ensures that the analytical measurement system is working within established limits. But it does not ensure consistency between operators, does not ensure good pre-analytical conditions, and does not ensure that the QA system supporting it is running correctly.
The right question is not "are our QC results always in range?" — but "is everything that needs to be protected, actually protected?"