Guide

Westgard rules for POCT, explained

In short

Westgard rules are decision rules applied to quality control results: some catch big one off errors, others catch slow drift. POCT services usually need simpler rule sets than main laboratories, chosen per device family. The two classic mistakes are using a single loose rule everywhere, which misses drift, and adopting a full multirule scheme everywhere, which rejects so many good runs that staff stop believing QC.

Why rules at all

A QC value can be inside the manufacturer's range and still be telling you something is wrong: six results in a row creeping upward is a story, not a coincidence. Westgard rules formalise that judgement so any trained operator makes the same decision.

The rules that matter for POCT

1_3s: one control more than three standard deviations from target. Almost always a genuine rejection: a big random error or a wrong control.
1_2s: one control beyond two standard deviations. As a rejection rule it generates false alarms; better used as a warning that triggers a look, not an automatic stop.
2_2s: two consecutive controls beyond two standard deviations on the same side. Systematic error: think reagent lot, calibration or storage.
R_4s: two controls four standard deviations apart in one run. Random error worth investigating.
10_x: ten consecutive results on the same side of the mean. Classic slow drift; on POCT devices this often flags a deteriorating sensor or an environmental change.

Choosing per device, not per textbook

Rule choice should follow the device's error behaviour and the clinical risk of the analyte. A hospital blood gas analyser with onboard QC deserves a multirule approach. A single strip based meter in a pharmacy may be well served by 1_3s with a drift check, run at a frequency you can actually sustain. Write the chosen rules into the SOP with the reasoning, because assessors ask why, not just what.

The over rejection trap

Every false rejection costs a control, ten minutes, and a little of the operator's faith in QC. Services that copy a full laboratory scheme onto simple devices train their staff, within months, to treat QC failure as noise. Fewer, better chosen rules, applied consistently, protect patients better than an impressive scheme nobody follows.

We teach this with real POCT data, including the awkward cases, in IQC and EQA Mastery for POCT.

Questions, answered

Which Westgard rules should a POCT service use?

It depends on the device family and clinical risk. A common pattern: 1-3s as the hard stop everywhere, a drift rule such as 10x, and 2-2s where two levels are run. Complex analysers justify fuller multirule schemes.

Is a 1-2s result a failure?

Best treated as a warning rather than an automatic rejection for most POCT devices, because using it as a hard stop produces frequent false alarms.

Do Westgard rules apply to unit use devices like strips and cartridges?

Yes, with adaptation. The rules read the QC data stream the same way, but frequency and rule choice reflect that each test uses a fresh unit rather than a shared reagent system.