Daily quality control for a small laboratory
Quality control is how a laboratory knows today’s results can be trusted before they go out. It does not need to be complicated, but it does need to be done the same way every day.
- RunTwo levels each test day
- PlotLevey-Jennings chart
- JudgeApply the rules
- ActFind the cause, then repeat
- RecordKeep charts and actions
How much quality control does a small laboratory need?
Run two levels of control on each day a test is performed, and one level every shift after that. Plot the values on a Levey-Jennings chart, judge them with a fixed set of rules, and release patient results only when the controls pass. NABL 112A sets this minimum for every laboratory, whatever its size, and also expects participation in an external quality assessment scheme.
Why re-running the control is not a fix
Many laboratories reject any run where a control falls outside two standard deviations, then simply repeat the control until it passes. With two controls, that rule alone falsely rejects about 9% of good runs, so repeating teaches staff to ignore it.
The Westgard multirule approach treats a single value beyond 2 SD as a warning, and rejects the run only when a further rule is broken. When one analyzer was assessed in an Indian tertiary laboratory, 5 of 21 tests performed below three sigma, which shows that some tests need closer control than others.
The Westgard rules in plain words
| Rule | What it means | Action | Likely error |
|---|---|---|---|
| 1-2s | One control beyond 2 SD | Warning; check the other rules | |
| 1-3s | One control beyond 3 SD | Reject | Random |
| 2-2s | Two in a row beyond 2 SD on the same side | Reject | Systematic |
| R-4s | Two controls in one run more than 4 SD apart | Reject | Random |
| 4-1s | Four in a row beyond 1 SD on the same side | Reject | Systematic |
| 10x | Ten in a row on the same side of the mean | Reject | Systematic |
Your laboratory director decides which rules apply to each test.
What to do when a control fails
Stop reporting
Hold patient results for that test.
Check the basics
Control expiry, storage, mixing and whether the right level was loaded.
Check the reagent
Lot change, on-board time and calibration status.
Check the analyzer
Maintenance due, probes, lamp and error log.
Fix, then repeat
Correct the cause, recalibrate if needed, and run both levels again.
Record it
What failed, what was done and who released the run.
How Milap helps
QC set-up at installation
Controls are run and ranges entered when the analyzer is commissioned.
Operator training
Running controls, reading the chart and what to do on a failure.
Preventive maintenance
Scheduled visits that keep performance from drifting.
Controls and calibrators
Supplied for the systems we install.
Sources
NABL 112A, Specific criteria for accreditation of medical laboratories.
Westgard QC, Westgard rules and multirule and false rejection.
Sigma metrics study, Indian Journal of Clinical Biochemistry.
Frequently asked questions
Yes. NABL 112A asks for two levels on the day of testing irrespective of the size of the laboratory.
No specific rule set is mandated. You need a documented procedure for judging controls, and Westgard rules are the most widely used.
A chart of control values over time with lines at the mean and at 1, 2 and 3 standard deviations, which makes shifts and trends visible.
NABL 112A sets retention periods; keep daily control values and charts at least until your next assessment.
An outside scheme sends unknown samples and compares your results with other laboratories. NABL expects participation in each discipline.
Want your team confident with daily QC?
Ask us about training and maintenance for your analyzers.