R-4s, 10x and finding the cause
14 min
Skills you'll build3 skills
- Recognize an R-4s violation within one run
- Recognize a 10x violation in consecutive same-side controls
- Match a QC shift, trend, or random error to its likely causes
At the bench: Review QC before releasing results
Try first
Get the idea
Two more rules complete the usual Westgard multirule set, and each detects a different type of error.
R-4s within one run
- The R-4s rule is met when one control is above +2 SD and another in the same run is below −2 SD, a spread of more than 4 SD.
- Controls measured together share the same reagent and calibration. When they move in opposite directions in the same run, the cause is random error.
- R-4s applies within one run. Pairing results from different runs includes between-run variation in the comparison.1,2
The 10x rule
- The 10x rule is met when ten consecutive controls fall on the same side of the mean, even if every one of them is inside ±2 SD.
- A procedure in control gives about half its results on each side of the mean, so ten in a row on one side indicates a small systematic shift.
- Some QC plans use 8x or 9x for this check, and the plan specifies whether the count includes results across levels as well as across runs.1,3
Investigating the cause
When a rule rejects the run, the type of error it detects indicates where to look first.2,3
| Error type | Rules | Causes to check first |
|---|---|---|
| Random | 1-3s, R-4s | Bubbles, a clot or short sample in the probe, uneven pipetting, electrical noise |
| Systematic | 2-2s, 4-1s, 10x, a shift or trend | Reagent lot or deterioration, calibration, temperature, a failing lamp or electrode |
Corrective action and lookback
Rerunning the same control until it passes does not identify or correct the fault. Work in this order:2,4
- Correct the cause.
- Rerun the controls.
- Release patient results only after QC is acceptable.
- Review the patient results from the rejected run and every result reported since the last acceptable run. Repeat those the error could have changed.
A sequence rule such as 10x is met only after a shift has lasted several runs, so the review goes back to the run where the shift began.3
Follow the bench demonstration
Sources for this step
References
- Westgard JO. Westgard rules and multirules. Westgard QC. Accessed September 23, 2026.
- Westgard JO. QC: the multirule interpretation. Westgard QC. Accessed September 23, 2026.
- Clinical and Laboratory Standards Institute. Statistical Quality Control for Quantitative Measurement Procedures: Principles and Definitions. 4th ed. CLSI guideline C24Ed4E. Clinical and Laboratory Standards Institute; 2016.
- Centers for Medicare & Medicaid Services. 42 CFR §§493.1256 and 493.1282. Electronic Code of Federal Regulations. Accessed September 23, 2026.
Watch one
A BUN level 2 control has a mean of 50 mg/dL and an SD of 2 mg/dL. Its last ten runs, oldest first:
| Run | 1 | 2 | 3 | 4 | 5 | 6 | 7 | 8 | 9 | 10 |
|---|---|---|---|---|---|---|---|---|---|---|
| Result | 50.8 mg/dL | 51.5 mg/dL | 50.4 mg/dL | 52.2 mg/dL | 51.0 mg/dL | 51.8 mg/dL | 50.6 mg/dL | 52.4 mg/dL | 51.2 mg/dL | 51.6 mg/dL |
A new calibrator lot was put into use before run 1. What do you decide, and where do you look?
- Check 1-2s: the +2 SD limit is 54 mg/dL and the highest result is 52.4 mg/dL, so there is no warning.
A warning would send you to the other rules, so it is checked first.
- Check 4-1s: only runs 4 and 8 are beyond the +1 SD limit of 52 mg/dL, and they are not consecutive, so 4-1s is not met.
Results within ±2 SD can still violate the sequence rules.
- Check 10x: all ten results are above 50 mg/dL, so 10x is met and the run is rejected.
Ten results in a row on one side of the mean indicate a small shift.
- Name the error type: systematic, so bubbles and pipetting are not the first place to look.
Every sequence rule detects systematic error, which shifts every result in the same direction.
- Check the calibration first: the new calibrator lot was put into use just before run 1, when the results moved above the mean.
The shift began when something common to every result changed.
Your turn
Use it
Optional for this step.
Review missed questions
For reference
The rest of this step has a short briefing, a demonstration at the bench, 3 practice problems and a case in the lab.
Free account: keep your progress and get review reminders. Have one? Sign in