Westgard Rules Explained for CLIA Labs

Westgard rules are statistical rules for deciding whether a QC run is acceptable. They are 1-2s (warning), 1-3s, 2-2s, R-4s, 4-1s and 10x, applied to control results plotted on a Levey-Jennings chart. CLIA does not require Westgard rules by name, but it does require defined QC acceptability criteria, documented QC, and corrective action whenever QC fails.

Last reviewed Oct 2026 · By Sound Lab Consultants CLIA team · General information, not legal advice.

Key facts and official sources

  • For quantitative tests, run two control levels at least once each day patient specimens are tested, unless an approved alternative such as IQCP applies. Source: 42 CFR 493.1256(d)(3)(i)

  • For quantitative controls, define statistical parameters (for example, mean and SD) for each lot. Source: §493.1256(d)(10)(i)

  • Control results must meet acceptability criteria before patient results are reported. Source: §493.1256(f)

  • When QC fails, evaluate patient results back to the last acceptable run, and document the corrective action. Source: §493.1282(b)(2)

  • Rule definitions: westgard.com, "Westgard Rules"

What are the Westgard rules?

  • 1-2s

    • Triggered when: One control is beyond ±2 SD

    • Usually signals: Warning. Inspect the run with the other rules

  • 1-3s

    • Triggered when: One control is beyond ±3 SD

    • Usually signals: Random error, so reject the run

  • 2-2s

    • Triggered when: Two consecutive controls beyond the same ±2 SD limit (same level across runs, or both levels in one run)

    • Usually signals: Systematic error, so reject

  • R-4s

    • Triggered when: Within one run, one control is above +2 SD and another below −2 SD

    • Usually signals: Random error, so reject

  • 4-1s

    • Triggered when: Four consecutive controls beyond the same ±1 SD limit

    • Usually signals: Systematic error (shift), so reject or investigate according to your QC plan

  • 10x

    • Triggered when: Ten consecutive controls on the same side of the mean

    • Usually signals: Systematic error (shift or trend), so reject or investigate according to your QC plan

Your QC policy should state which rules you apply to each test, and whether 4-1s and 10x reject the run or trigger a review.

How do you build a Levey-Jennings chart?

  1. Set the mean and SD. Run the new control lot over time, typically 20 or more data points across multiple runs. Calculate the mean and SD. You can use the manufacturer's assayed values only if they apply to your method and instrument and you verify them (§493.1256(d)(10)(ii)).

  2. Draw the limits. Plot horizontal lines at the mean and at ±1, ±2 and ±3 SD.

  3. Plot every result in date order, keeping each control level on its own chart.

  4. Apply your rules before releasing patient results.

  5. Review regularly. The TC or TS reviews and signs the charts, at least monthly is common practice. Watch for shifts and trends.

  6. Reset for each new lot. Run the new lot alongside the old one before switching, then establish the new lot's own statistics.

What do you do when a Westgard rule is violated?

  • Stop reporting results from that test system.

  • Identify the error type: random (1-3s, R-4s) or systematic (2-2s, 4-1s, 10x).

  • Check the controls, reagents, calibration, maintenance and the instrument's flags.

  • Correct the problem and rerun the controls.

  • Evaluate patient results back to the last acceptable run, and correct any reported results that were affected.

  • Document every step. See our QC corrective action log guide.

Repeating controls until one passes, with no investigation, is a common survey finding.

Does CLIA require Westgard rules?

No. CLIA requires the lab to establish or verify acceptability criteria, follow them, document QC and correct failures. Westgard multirules are a widely used way to meet that. Accreditors and lab policies may specify rules.

Frequently asked questions

Is the 1-2s rule a rejection rule?

In the classic multirule scheme, 1-2s is a warning that prompts checking the other rules, not an automatic rejection.

How many data points are needed to set a new mean?

A common practice is at least 20 data points across multiple runs. Follow your lab's written QC policy.

Can we use the manufacturer's control ranges?

You can use an assayed control's stated value only if it applies to your methodology and instrument and you verify it. Many labs narrow the range using their own data.

Do Westgard rules apply to qualitative tests?

No. They are for quantitative controls. Qualitative tests use positive and negative controls.

Need help with your QC program?

Sound Lab Consultants writes QC policies, sets up Levey-Jennings review and fixes QC survey findings. myLabCompliance.io tracks QC with Levey-Jennings charts and Westgard rules.

Book a free consultation