Why Quality Control Matters in the Diagnostic Laboratory

Every result that leaves a diagnostic laboratory carries clinical weight. A haemoglobin of 92 g/L, a sodium of 128 mmol/L, a TSH just outside the reference interval — clinicians act on these numbers, and patients feel the consequences. Quality control (QC) is the routine, internal process that demonstrates those numbers are fit for purpose before they are reported.

It helps to keep three ideas separate. Quality control is the daily testing of materials with known values alongside patient samples. Quality assurance is the wider system — standard operating procedures, training, equipment maintenance, audits. External quality assessment is the periodic, blinded comparison of your results with other laboratories. QC is your early warning system; the other two tell you whether the whole system is holding together.

Choosing and Handling Control Materials

Control materials should behave as closely as possible to patient samples. That sounds obvious, but it drives most of the practical decisions.

• Use at least two levels. A normal and an abnormal level — covering the clinically important decision points — will catch errors that a single mid-range control misses.
• Match the matrix to the method. A control validated for serum may not suit a plasma or whole-blood assay, and lyophilised controls can behave differently from liquid-stable ones.
• Reconstitute carefully. Use a volumetric pipette, allow the vial to stand as the manufacturer specifies, and mix by gentle inversion rather than vigorous shaking.
• Aliquot and freeze. Repeated freeze–thaw cycles degrade many analytes. Decant single-use aliquots and label them with the lot number and date.
• Respect expiry dates and storage conditions. A control stored at the wrong temperature may still produce a plausible number, which is exactly what makes it dangerous.
• Treat controls like patient samples. Same analyser, same handling, same staff, same conditions. If you only use them on a "good" day, you are testing your optimism rather than your method.

Charting Results and Applying Statistical Rules

Once a method is stable, establish a mean and standard deviation from at least 20 separate runs. Plot each day's control result on a Levey–Jennings chart, with the mean as the centre line and lines at ±1, ±2 and ±3 standard deviations. The chart turns a string of numbers into a picture, and pictures reveal shifts and trends long before a single point breaks a rule.

Most UK laboratories use a multirule approach, widely known as Westgard rules, to decide when to accept or reject a run. In practice you will meet:

• 1-2s — a warning rule. One control outside ±2 SD prompts inspection, not necessarily rejection.
• 1-3s — one control outside ±3 SD. Reject the run; look for random error.
• 2-2s — two consecutive controls outside the same ±2 SD limit. Suggests systematic error, such as a calibration drift.
• R-4s — the difference between two controls within a run exceeds 4 SD. Points to random error or a handling problem.
• 4-1s and 10x — four consecutive results beyond 1 SD, or ten on one side of the mean. These flag bias developing quietly.

Keep charts visible on the bench or dashboard, not buried in a folder. A trend is easier to act on at day three than at day ten.

What To Do When Quality Control Fails

Stop. Do not report patient results from that run until the problem is understood. Repeating the control until it passes, then carrying on, is not troubleshooting — it is wishful thinking with a paper trail.

• Check the obvious first. Expired or incorrectly stored reagents, wrong control level, air bubbles, a clot in the sample probe, or a control vial left at room temperature overnight account for a surprising share of failures.
• Repeat with a fresh aliquot of the same control. If it passes, suspect the aliquot rather than the method.
• Open a new vial if the repeat fails, particularly if the current vial is near expiry or has been in use for some time.
• Review the analytical system. Calibration status, reagent lot change, maintenance due, lamp hours, temperature stability.
• Look back at patient results since the last acceptable QC. If a genuine failure is confirmed, consider whether results need withholding, repeating or amending.
• Document everything. What failed, what you checked, what you changed, and the outcome. Records are the evidence that your corrective action was reasoned.
• Escalate when unresolved. Involve the section lead, quality manager or clinical scientist rather than releasing results on hope.

Meeting UK Accreditation Expectations

Laboratories accredited to ISO 15189 — the standard against which the national accreditation body assesses medical laboratories — are expected to demonstrate QC as part of a managed system, not as an informal habit. In practice, assessors will look for a written QC policy with defined acceptance criteria, documented procedures for each analyte, and evidence that staff follow them.

They will also expect records: control lot numbers, means and standard deviations, Levey–Jennings charts, failure investigations and corrective actions, retained for the period your quality manual specifies. Competency assessment matters too — a scientist should be able to explain, in their own words, why a run was rejected and what they did next. Participation in external quality assessment schemes, reviewed with action taken on poor performance, completes the picture.

Making Quality Control Part of Everyday Practice

The best QC cultures are not built on fear of inspection. They are built on small, consistent habits: checking controls before the first patient sample, recording deviations honestly, talking through failures at handover, and treating a rejected run as useful information rather than a nuisance.

Train new staff early and let them troubleshoot under supervision, so the reasoning becomes instinctive. Review your QC data monthly for trends that daily rules miss. And remember the point of all of it — a number on a report is only as good as the evidence behind it, and that evidence is generated quietly, every morning, one control vial at a time.