Chapter 1 · Analyser QC and calibration in biochemistry

Clinical biochemistry relies heavily on automated analysers. Each analyser must have a documented QC schedule running at least two levels (normal and abnormal) at the start of each shift and at defined intervals. Calibration must be performed with reagent lot changes, after maintenance, and at the manufacturer’s recommended intervals. Westgard rules must be applied. Delta checks are particularly important in biochemistry: electrolytes, urea, and creatinine rarely change abruptly between consecutive results in a stable patient.

Chapter 1 · Practical example

Haemolysis interference — K+ and the falsely elevated result

Haemolysis releases intracellular potassium into serum, falsely elevating K+ results. Recognising this interference and acting on it protects patients from unnecessary treatment.

▲ The situation

A haemolysed serum sample from the emergency department shows K+ 7.4 mmol/L. The previous K+ on this patient, taken two days ago, was 4.1 mmol/L. The haemolysis index is critically elevated at 5+.

Reference: ISO 15189:2022 Clause 7.3.7; Clinical Biochemistry Reference

✓
Good practice

The analyst checks the haemolysis interference threshold for potassium in her method — exceeded. She rejects K+ with a specific comment: K+ result invalid due to marked haemolysis (haemolysis index 5+). Haemolysis falsely elevates potassium. Previous K+ 4.1 mmol/L (two days ago). Recollection recommended urgently. She contacts the ED directly to explain, reports the troponin and sodium (not significantly affected at this haemolysis level) with a haemolysis comment.

✗
Poor practice

All results including K+ 7.4 are released with a generic comment haemolysis noted. The ED team initiates urgent hyperkalaemia treatment — calcium gluconate, insulin-dextrose, bicarbonate. Repeat sample collected an hour later shows K+ 4.0 mmol/L. The patient received unnecessary treatment carrying its own risk. The laboratory had the information to prevent this.

● What this means for your practice

Haemolysis is an analyte-specific interference, not a global sample problem. Know which analytes in your repertoire are affected and at what haemolysis index. Release unaffected analytes, reject affected ones, and make the clinical implications explicit in your comment.