10.3 Quality Control

10.3 Quality Control

Quick Answer: Run QC on schedule with the correct levels and lots; accept only results inside the lab’s acceptable ranges; if out of control, hold patients, troubleshoot step-by-step (rerun → new control → reagent/instrument checks → escalate), and document everything. Know simple Westgard-style ideas (shift vs trend, random vs systematic) as concepts that drive actions—not as a statistics course.

Quality control is how the laboratory proves the analytical system is stable before trusting patient numbers. For Domain III, the MLA must be able to perform QC, spot failures, take first-line corrective steps, and know when to stop and call the MLS. You are not expected to redesign control limits or compute complex multi-rule statistics from scratch on the exam, but you are expected to behave correctly when a control is out.

Why QC Exists

Patient specimens are unknown. Controls are materials with expected values that travel through the same pipetting, reagents, incubation, and detection path as patients. When controls land inside limits, confidence rises that patient results from that run/system are analytically reliable. When controls fail, patient results for the affected tests are suspect until the system is restored and documented.

QC protects against:

• Degraded or wrong reagents
• Calibration drift
• Pipettor/probe problems
• Temperature failures
• Operator process errors
• Environmental issues (bubbles, clots, bad water)

Running QC per Schedule

Frequency principles (follow your SOP)

CLIA and manufacturers set minimums; labs often tighten them. Common patterns:

Never skip QC because the board is full of STAT samples. The correct sequence is: restore control → then patients (or use a backup method that is itself in control).

Levels and materials

• Many quantitative assays use two or three levels spanning the measuring range (e.g., low and high, or low/normal/high)
• Use the assigned lot and the software file that matches that lot’s target/ranges
• Bring controls to the correct temperature and mix gently; reconstitute exactly when lyophilized
• Do not use expired controls even if “they were fine yesterday”

Acceptable Ranges and Evaluation

Each control level has an acceptable range established by the lab (often mean ± 2 SD or manufacturer/lab-verified limits displayed in the LIS/middleware).

In control

• All required levels fall within their ranges (and any multi-rule policy is satisfied)
• No unresolved instrument flags on the QC cup
• Documentation complete
• Patient testing may proceed for those assays

Out of control

Any of the following (examples—SOP is authoritative):

• One or more levels outside the acceptable range
• A rule failure such as two consecutive levels beyond 2 SD on the same side (conceptual Westgard-type rules labs encode in software)
• QC not run when due (overdue = not acceptable to proceed)
• Obvious process error (wrong control in the cup)

If out of control: do not report patient results generated on the failed system/assay until corrective action succeeds and policy allows release (including possible patient repeat decisions by MLS).

Out-of-Control Recognition: Practical Signs

• Red flags or “fail” icons on the QC review screen
• Values printed outside the listed range on the package insert sheet used by the lab
• Sudden jump from a stable history (e.g., always ~100, today 145 with limit 120)
• Only one level failing vs both levels failing (pattern helps troubleshooting)
• Several assays failing together (points to shared reagent probe, water, or calibration event)

Train your eye to open the QC review before releasing a rack of outpatients out of habit.

Troubleshooting Steps (Entry-Level Sequence)

Use a calm, ordered approach. Random button-pressing wastes controls and time.

Step 1 — Pause and contain

• Stop releasing affected patient results
• Note which assays, levels, instrument, and lot failed
• Alert the bench MLS early if the failure is widespread or STAT work is stacking

Step 2 — Quick obvious checks

• Correct control material and level in the correct position?
• Control expired or past open dating?
• Control mixed and at proper temperature?
• Reagent onboard expired/empty/low?
• Recent calibration or maintenance unfinished?
• Air bubbles, clot flags, or short sample on the QC cup?

Many “QC failures” are wrong cup, wrong level, or expired vial—fix those first.

Step 3 — Rerun QC

• Rerun the failed level(s) once with careful technique if SOP allows a single rerun
• If rerun is in → investigate whether the first failure was random error; document; follow rules for whether patients run between the failure and the pass need review
• If rerun is still out → do not keep rerunning indefinitely

Step 4 — New control material

• Open a fresh vial/bottle of the same lot (or next steps per SOP if vial integrity is questionable)
• Reconstitute carefully if needed
• Rerun required levels

Step 5 — Reagent and system checks

• Replace suspect reagent pack/lot if indicated
• Verify water, diluent, waste, temperatures
• Check for probe clogs or maintenance due items
• Review whether calibration is due or was just rejected

Step 6 — Escalate

If still out of control:

• Notify MLS/supervisor/technical specialist
• Consider backup instrument or send-out per downtime procedures
• MLS may perform advanced troubleshooting, service calls, or calibration review
• Document downtime and patient impact

Documentation Requirements

If it is not documented, inspectors and future troubleshooters assume it did not happen.

Record:

• Date/time
• Instrument and assay
• Control lot numbers and levels
• Results and pass/fail
• Corrective actions (rerun, new bottle, reagent change, service)
• Who performed QC and who reviewed failures
• Patient result disposition decisions when applicable (usually MLS)

Electronic QC modules still need honest comments—do not clear a failure without a real fix.

Westgard-Style Awareness Without Stats Mastery

Levey-Jennings charts plot control results over time. Westgard multi-rules are combinations of rules labs use to detect error efficiently. For the MLA exam, master the ideas:

Random error

Unpredictable scatter—bubbles, occasional clots, brief mix-ups. May produce a single outlier. Action: careful rerun, technique check, document.

Systematic error

Consistent bias in one direction—bad calibration, degraded reagent, wrong temperature set point. May show as a shift (sudden sustained move) or trend (gradual climb or fall). Action: do not ignore “almost out” drift; escalate before every level is wildly out.

Common conceptual rules you may see named

• $1_{2s}$: one point beyond 2 SD — often a warning to inspect, not always automatic reject (lab-dependent)
• $1_{3s}$: one point beyond 3 SD — typically reject
• $2_{2s}$: two consecutive points beyond 2 SD on the same side — reject (systematic concern)
• $R_{4s}$: two points in a run differ by more than 4 SD — random error concern

You do not need to compute SDs by hand under time pressure. You do need to respect the software’s fail flags and the SOP’s response tree.

Shift vs trend (visual)

• Shift: many points suddenly center at a new mean after a reagent or calibration change
• Trend: points slowly march up or down across days

Both can precede a hard out-of-range event. Good MLAs mention developing trends to the MLS instead of waiting for a red fail on a full outpatient morning.

Patient Results and QC Failures

Core rule: Do not release patient results for assays that are out of control.

MLS decisions may include:

• Repeating patients after QC is restored
• Evaluating results reported after the last in-control QC and before the failure
• Using a backup method
• Delaying non-critical tests during downtime

The MLA supports by holding releases, retrieving specimens, preparing repeats, and documenting timeline facts—not by deciding clinical acceptability alone.

Integrated Scenarios

Scenario A: Level 1 sodium control is 5 mmol/L above the upper limit; Level 2 is in range. You verify the correct Level 1 vial, mix, and rerun once—still high. You open a new Level 1 vial—still high. Next: check ISE reagents/calibration status and escalate; hold sodium patient releases.

Scenario B: Both levels of glucose fail high immediately after a new reagent pack is installed. Suspect pack seating, pack integrity, or need for calibration/QC sequence per SOP. Do not send a full diabetic clinic rack “because they are waiting.”

Scenario C: QC was due at 0600; it is 10:00 and ambulatory patients were run. Discovery of missed QC is a quality event. Stop, run QC now, notify MLS for impact assessment of results reported without current QC—not quietly backfill timestamps.

Scenario D: A Levey-Jennings display shows glucose Level 2 climbing for five days but still inside limits. Tell the MLS; check reagent age/lot; do not wait for an automatic $1_{3s}$ failure during a holiday weekend.

Link to Practice and Related Skills

QC links reagents (10.1), instruments (10.2), waived/POC QC, critical values, and laboratory QA/QI in Domain IV. Build judgment with free questions at /practice/ascp-mla.

Key Takeaways

• Run the right controls on the required schedule with in-date, properly handled materials.
• Evaluate against acceptable ranges; out-of-control means hold affected patient results.
• Troubleshoot in order: obvious checks → rerun → new control → reagent/system → escalate.
• Document lots, results, failures, and corrective actions completely.
• Use Westgard/Levey-Jennings concepts to understand random vs systematic problems and to take timely MLA actions.