SN4 SHORT NOTE 5M ★ High-yield

Describe quality control in the clinical biochemistry laboratory.

Biochemistry · 1st Professional — Pre-Clinical

Definition

Quality control = the system of procedures that ensures laboratory results are reliable, reproducible and clinically usable.

Term	Meaning
Accuracy	Closeness to the true value; affected by systematic error (bias)
Precision	Reproducibility on repeated measurement; affected by random error
Sensitivity	Ability to detect true positives; a sensitive test is good for ruling out
Specificity	Ability to identify true negatives; a specific test is good for ruling IN
Coefficient of variation	(SD ÷ mean) × 100 — the usual measure of precision

Mnemonic — SnNOut (a Sensitive test, if Negative, rules Out) and SpPIn (a Specific test, if Positive, rules In)
A method can be precise but inaccurate — consistently wrong. Both are needed

Types of Quality Control

Type	Description	Purpose
Internal (IQC)	Control sera of known value run with each batch, daily	Monitors precision day to day
External (EQAS)	Unknown samples sent by a central agency and compared between laboratories	Monitors accuracy against peers

The Levey–jennings Chart

Control values are plotted daily against the mean ± 1, 2 and 3 standard deviations
95% of results should fall within ± 2 SD
Patterns are interpreted by the Westgard rules

Westgard rule	Meaning	Suggests
12s	One value beyond 2 SD	Warning only — expected in 1 of 20 runs
13s	One value beyond 3 SD	Reject — random error
22s	Two consecutive beyond the same 2 SD limit	Reject — systematic error
R4s	Two values differing by more than 4 SD	Random error
10x	Ten consecutive on the same side of the mean	Shift — new reagent lot, recalibration
7T	Seven consecutive rising or falling	Trend — deteriorating reagent, ageing lamp, electrode drift

CLINICAL PEARL

A shift is abrupt; a trend is gradual. The distinction points to different causes — a shift usually follows a discrete event such as a new reagent lot or recalibration, while a trend suggests something deteriorating steadily.

Sources of Error

Phase	Share of errors	Examples
Pre-analytical	60–70% — the largest	Wrong patient or label, haemolysis, wrong tube, tourniquet too long, delay in transport, non-fasting sample
Analytical	10–15%	Instrument fault, reagent deterioration, calibration error, interference
Post-analytical	15–20%	Transcription error, delayed reporting, failure to communicate a critical value

Most laboratory error occurs before the sample reaches the analyser — which is why sample collection deserves as much attention as the assay
Haemolysis falsely raises potassium, LDH, ast and phosphate — the commonest single interference

Applied Aspects

Critical (panic) values must be telephoned immediately — potassium above 6.5, glucose below 40, calcium above 13, and a positive blood culture
Always question a result that does not fit the patient; repeat before acting on a surprising value
Reference ranges are population-specific, defined as the central 95%, which means 1 in 20 healthy people falls outside by definition — the reason a panel of 20 tests so often shows one "abnormal" result

All questions in Clinical & Applied Biochemistry

Every Biochemistry chapter →