Category: Data-Driven Troubleshooting | Estimated reading time: 9 minutes
After a Reject, What Actually Happens?
Preliminary testing establishes the foundation. Levey-Jennings charts visualize the data. Westgard rules provide decision criteria. TE vs TEa and Sigma metrics evaluate performance quantitatively.
But all of that only holds meaning if one thing happens afterward: appropriate action.
A QC reject without proper investigation is a wasted QC reject—consuming time, reagents, and energy, but producing no useful information or real improvement. Even worse: if every reject is met only with "repeat until it passes," we are building a culture that normalizes problems rather than solving them.
This article discusses what happens after a QC reject—and why the way we respond to rejects is the most honest reflection of a laboratory's quality system maturity.
Starting Point: Diagnosis Before Action
When a QC result is rejected, the most common first response in many laboratories is: repeat. This is not absolutely wrong—but it is incomplete if done without the preceding question:
"Random error or systematic error?"
This question is not a formality—it is a diagnosis. Just like in the medical field, the appropriate action can only be chosen after a correct diagnosis.
Repeating the measurement is an appropriate response for random error, because its nature is conditional and it may not recur. But repeating a measurement for a systematic error will only yield another rejected result—because the cause lies in the system, not in the individual measurement.
Left Branch: Random Error (1₃s / R₄s)
When the triggered rule is 1₃s (one point outside ±3SD) or R₄s (range of two consecutive points exceeding 4SD), the diagnosis points toward random error. There are four areas of investigation:
Check the sample (control material)
- Has the vial been open for too long? Control material open for longer than the recommended time (usually stated in the kit insert) can undergo evaporation or degradation that alters its values.
- Was the preparation (reconstitution) performed correctly? For lyophilized control materials, errors in solvent volume or homogenization time can produce concentrations different from expected.
- Is there any contamination in the vial?
Check the condition of the control material
- Is the storage correct—proper temperature, not exposed to direct light, not re-frozen after thawing?
- Are there suspicious visual changes—different color, sediments, unusual bubbles?
Check operator pipetting technique Dramatic random errors often stem from technical errors in a specific measurement: air bubbles in the pipette tip, a tip that is not attached perfectly, sample not mixed homogenously before aspiration, or dead volume that was overlooked. This is an area often left unexamined because it involves evaluating the operator—yet it is most frequently the source of dramatic random errors.
Check the instrument condition at the time
- Are there any alarms or error messages from the instrument that might have been missed?
- Were there changes in physical conditions just before the measurement—unstable voltage, vibrations, sudden temperature changes?
- Is the instrument probe clean and free of clogs?
Right Branch: Systematic Error (2₂s / 4₁s / 10x)
When the triggered rule is 2₂s, 4₁s, or 10x, the diagnosis points toward systematic error—a condition more serious than random error, as it means something has changed permanently or semi-permanently within the system. Repeating the measurement will not solve this problem. There are four areas of investigation:
Check instrument calibration Calibration is the most common source of systematic error. Check: when the last calibration was performed and whether it has passed the recommended interval; whether there are conditions that could cause drift faster than usual (intensive usage, reagent changes, environmental conditions); whether the calibrators used are still valid (not expired, stored correctly).
In many cases, recalibration is the most effective first step for a systematic error—but recalibration without understanding why the drift occurred only solves the problem temporarily, without preventing its recurrence.
Check reagent lot (new vs. old)
Switching reagent lots is one of the most frequently overlooked causes of systematic error, especially if done without proper verification. Check: has there been a lot change in the recent period; has the new lot been verified with a mini preliminary test before being used for patients; is the new lot's characteristic consistent with the previous lot? (This is the reason why preliminary testing needs to be repeated when changing reagent lots.)
Check reagent storage temperature Reagents exposed to improper temperatures—too cold or too hot—can undergo degradation that alters their reactivity. Check: has there been a cooling system disturbance recently; has any reagent accidentally been exposed to extreme temperatures (e.g., left outside the refrigerator too long during preparation); does refrigerator temperature monitoring show unusual fluctuations.
Check recent instrument maintenance Incomplete or irregular maintenance can cause gradual changes in instrument performance. Check: when the last maintenance was performed and whether all steps were followed correctly; are there components that have exceeded their replacement interval (probe, tubing, lamp, filter); is there a history of technical issues in the recent period that have not been fully resolved.
End Point: Documentation → Action → QC Re-verification
After investigation and corrective actions are taken, there are three closing steps that must not be skipped.
Documentation It is not an administrative formality—it is institutional memory that allows the laboratory to learn from every QC reject incident. Good documentation records: the Westgard rule triggered, the investigation results (what was found), the actions taken, who performed the investigation and action, and the date and time of the incident. Without good documentation, recurring problem patterns will never be identified—the same problem could keep happening without realizing it is the tenth time in six months.
Action Actions must match the diagnosis: for a random error with an identified cause, correct that specific cause; for a systematic error, correct the systemic source (recalibrate, change reagent, perform maintenance); if the cause is not identified, escalate to a supervisor or the instrument's technical support.
What must not be done: releasing patient results before QC returns to acceptable status after corrective action. This is a fundamental principle that cannot be compromised—regardless of time pressure or work volume.
QC Re-verification After corrective action, QC must be repeated to verify that the problem has truly been resolved. A QC result that returns to acceptable status after corrective action confirms that the action taken was on target. If the QC is still rejected after corrective action, the investigation continues—this cycle must not be rushed by skipping any steps.
Culture Behind Procedures
There is one thing that cannot be depicted in any flowchart, but determines whether this system truly works: the courage to stop.
In a busy laboratory, with long sample queues and time pressure from clinicians, the decision to hold all results until a QC reject is resolved requires real courage. There is pressure—explicit or implicit—to "keep going."
But every time we release results without valid QC, we are borrowing trust from patients and doctors without any guarantee that such trust is deserved.
Plebani (2017), in a broader context of laboratory quality, asserts that a patient safety culture in a laboratory is not built by systems or regulations alone—it is built by individual decisions made correctly, one after another, even when no one is watching.
Closing
This flow builds one cohesive sequence: new control material arrives → preliminary testing establishes TV and SD → daily data plotted on LJ charts → Westgard rules provide decision criteria → TE vs TEa and Sigma metrics evaluate performance quantitatively → and when a QC reject occurs, proper investigation determines the right action.
Every step in this flow is connected to the next. This system only works well if all steps are carried out—not cherry-picked based on convenience or time pressure.