Chapter 1 · Internal quality control strategy

Internal QC (IQC) monitors the ongoing performance of examination processes. QC materials must be treated like patient samples. QC must be run before patient testing begins and at regular intervals during the run. Levey-Jennings charts must be maintained and reviewed. QC targets and limits must be established by the laboratory, not taken directly from the manufacturer.

Chapter 1 · Practical example

Designing an IQC strategy — two levels, every analytical run

IQC must be designed to detect both systematic and random errors. Two levels of control — one within the reference interval, one outside — are the minimum for clinical laboratory testing.

▲ The situation

A small clinical laboratory currently runs a single normal-level QC material once per shift for its biochemistry panel. The quality manager is reviewing whether this meets ISO 15189:2022 requirements.

Reference: ISO 15189:2022 Clause 7.3.7

✓
Good practice

The quality manager reviews the evidence: single-level QC detects bias at the control concentration but cannot detect changes in imprecision that may affect results at other concentrations — particularly at medically critical concentrations above or below the reference interval. She implements two-level QC — one within the reference interval, one in the medically critical zone — run at the start of every analytical run. The additional cost is minimal; the detection capability is substantially improved.

✗
Poor practice

The single-level QC continues because the laboratory manager says it has always worked and adding a second level would be expensive. During an accreditation assessment, the assessor reviews the IQC strategy and asks how the laboratory detects analytical errors at concentrations outside the normal range. The answer is that it cannot. This is a finding — the IQC strategy is insufficient for the claims of clinical performance.

● What this means for your practice

IQC strategy is a deliberate design decision, not a tradition. Two levels detect more error types than one. The clinical concentrations that matter most — the ones at decision thresholds — must be covered by at least one QC concentration.