Every medical laboratory professional knows the quiet confidence that comes with seeing QC results fall within acceptable limits. Westgard rules satisfied. Controls in range. Green across the board. But here’s an uncomfortable truth: your QC can pass and your patient results can still be wrong.
Understanding why this happens is one of the most important and underappreciated competencies in laboratory medicine.
The Illusion of the Acceptable Range
Quality control is a sampling exercise. We run two or three control levels, once or twice a shift, and use those handful of data points to make inferences about the entire analytical run. Statistically, this means that errors of moderate magnitude can slip through without triggering a single rule violation. This is not a flaw in QC design — it is a mathematical reality called the probability of error detection, and it is rarely 100%.
Errors That Can Hide in Plain Sight
1. Matrix Effects and Commutability Failures
Commercial controls are not patient samples. Their protein composition, viscosity, and analyte binding differ from real specimens. An analyzer may behave differently with control material than with actual blood or urine, meaning the control “passes” while patient samples carry a systematic bias that goes entirely undetected.
2. Lot-to-Lot Reagent Variation
When a new reagent lot is introduced without adequate lot verification, a shift in patient results can occur. If the new controls came from the same reagent lot, they absorb the same shift and nothing looks out of place. The error and the yardstick moved together.
3. Errors Outside the Measuring Interval
Standard QC concentrations are typically placed at medical decision points in the mid-range. Results at the extreme low or high end of the analytical measurement range may carry significant imprecision or bias that your mid-level controls will never detect.
4. Carry-Over and Sample Interaction
High-concentration specimens can contaminate the next sample in the queue. QC material, run in a controlled sequence, rarely encounters this real-world phenomenon. The patient with a critically elevated result may silently elevate their neighbour’s.
5. Pre-Analytical Variation
No QC monitors what happens before the sample reaches the analyzer. Haemolysis, incorrect tube selection, delayed centrifugation, mislabeling, or prolonged sample transit are invisible to even the most rigorous analytical QC program. Studies consistently show that 60–70% of laboratory errors originate pre-analytically.
6. Sigma Shifting at the Population Level
A method may perform well on average, keeping QC in range, while exhibiting increased imprecision for specific patient populations — such as those on interfering medications, with unusual protein levels, or with rare genetic variants affecting assay binding.
What Can We Do About It?
- Run patient-based QC (PBQC) alongside traditional controls. Moving averages of patient results are independent of control matrix and detect shifts that commercial QC misses.
- Perform robust lot verification before switching reagents, calibrators, or even control lots — using patient samples, not just the new controls.
- Monitor delta checks rigorously. Unexpected changes in serial patient results are often the earliest signal of an analytical problem.
- Review outlier rates and flagging patterns periodically. A subtle increase in instrument flags or repeat requests is a soft signal worth investigating.
- Never treat pre-analytical quality as someone else’s problem. Train, audit, and advocate across the entire testing cycle.
Acceptable QC is necessary, but it is not sufficient. The most dangerous errors in a laboratory are not the ones that trigger alarms. They are the ones that pass quietly, result by result, until a clinician notices something doesn’t fit the clinical picture.
Our patients deserve a QC mindset that goes beyond the control chart.