Skip to content
LatticeMLS, Home

Add your earlier progress to your account?

From before you signed in, this browser has:

Which plan do you want to keep?

You changed your plan before signing in, and those changes don't match the plan saved on your account.

SettingSaved planPlan changes

Study progress is waiting to be saved

Reading a Levey-Jennings chart

14 min

Skills you'll build4 skills
  • Recognize a shift, trend, or random scatter on a Levey-Jennings chart
  • Tell a 1-2s warning from a 1-3s rejection
  • Tell a 2-2s violation from a 1-3s violation
  • Tell a 2-2s violation from a 4-1s violation

At the bench: Review QC before releasing results

Try first

Try first

A calcium level 2 control has a mean of 11.6 mg/dL and an SD of 0.2 mg/dL. These are its last eight runs:

Run12345678
Result11.5 mg/dL11.7 mg/dL11.5 mg/dL11.6 mg/dL11.9 mg/dL11.8 mg/dL11.9 mg/dL11.9 mg/dL

What pattern does the chart show?

Keep going to see the reasoning.

Right. Keep going to follow the reasoning.

Keep going to see the reasoning.

Keep going to see the reasoning.

Get the idea

The chart

A Levey-Jennings chart plots each control result against the run in which it was measured.

  • Lines mark the mean and 1, 2 and 3 SD on either side of it, drawn from that control lot's own mean and SD.
  • Each control level has its own chart, so a result is read as its distance from its own mean in SD.1

Random and systematic error

Two kinds of error show on the chart:1

  • Random error scatters results unpredictably to both sides, as a bubble or an uneven pipette delivery would.
  • Systematic error pushes results one way. It appears as a shift, a sudden move to a new level that stays, or a trend, a steady drift in one direction. Failing reagent, a calibration problem or a temperature change can cause either one.

The Westgard rules

The Westgard rules define when those patterns accept or reject a run. Every selected rule is checked with each new result, together with the earlier results the rule needs.2,3

RuleMet whenUsual meaning
1-2sOne result beyond 2 SDWarning: check the other rules
1-3sOne result beyond 3 SDReject: random or systematic error
2-2sTwo consecutive results on the same side beyond 2 SDReject: systematic error
4-1sFour consecutive results on the same side beyond 1 SDReject: small systematic error

Three pairs of rules are often confused:2,3

  • The 1-2s and 1-3s rules. A result between 2 and 3 SD is a warning. A result beyond 3 SD rejects the run regardless of the other level.
  • The 2-2s and 1-3s rules. A 2-2s violation needs two results on the same side beyond 2 SD, and neither has to reach 3 SD. The two results can be both levels in one run or one level in two runs in a row.
  • The 4-1s and 2-2s rules. A 4-1s violation needs four results on the same side beyond 1 SD, and none has to reach 2 SD, so a run can meet 4-1s with no 1-2s warning.

When a rejection rule is met, patient results from the run stay pending release while the cause is found and corrected.3

Follow the bench demonstration

Sources for this step
References
  1. 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.
  2. Westgard JO. Westgard rules and multirules. Westgard QC. Accessed September 23, 2026.
  3. Westgard JO. QC: the multirule interpretation. Westgard QC. Accessed September 23, 2026.

Watch one

Glucose runs two control levels. Level 1 has a mean of 95 mg/dL and an SD of 3 mg/dL. Level 2 has a mean of 285 mg/dL and an SD of 8 mg/dL. Both levels were near their means in the last run. Today's run gives these results:

ControlMeanSDToday
Level 195 mg/dL3 mg/dL101.9 mg/dL
Level 2285 mg/dL8 mg/dL303 mg/dL

Do you accept the run?

  1. Convert each result to SD from its mean: (101.9 − 95) ÷ 3 = +2.3 SD and (303 − 285) ÷ 8 = +2.25 SD.

    Each level is read against its own mean and SD, so the two can be compared on one scale.

  2. Check 1-3s: neither result goes beyond 3 SD, so 1-3s is not met.

    A single result beyond 3 SD rejects the run, so it is checked first.

  3. Check 1-2s: both results are beyond +2 SD, so the warning is raised.

    A result beyond 2 SD is a warning that sends you to the other rules.

  4. Check 2-2s: both levels are above +2 SD in the same run, so 2-2s is met.

    Two results on the same side beyond 2 SD, both levels in one run, meet 2-2s even though neither is extreme.

  5. Name the error type: 2-2s signals systematic error, so the investigation looks at reagent, calibration and temperature.

    Both levels moved the same way, which suggests a cause common to every specimen in the run.

Reject the run for 2-2s and keep its patient results pending release while the systematic error is found.

Practice this skill

Your turn

Problem 1 of 3 · Hints available

Potassium level 1 has a mean of 3.20 mmol/L and an SD of 0.08 mmol/L. Level 2 has a mean of 6.20 mmol/L and an SD of 0.14 mmol/L. Today's run:

ControlMeanSDToday
Level 13.20 mmol/L0.08 mmol/L3.48 mmol/L
Level 26.20 mmol/L0.14 mmol/L6.25 mmol/L

What does the run show?

Level 1 is (3.48 − 3.20) ÷ 0.08 = +3.5 SD from its mean, beyond the 3 SD limit. One result there rejects the run, regardless of level 2.

The warning covers results between 2 and 3 SD. Level 1 is +3.5 SD, so the run is rejected even though level 2 sits near its mean.

Treated a control beyond 3 SD as a warning

One control result beyond 3 SD violates the 1-3s rule, and the run is rejected regardless of the other level. The 1-2s warning applies to results between 2 and 3 SD. Treating a 3 SD result as a warning releases patient results from a run with large random or systematic error.

2-2s needs two results on the same side beyond 2 SD. Level 2 is only +0.4 SD, so only one result is out, and it is beyond 3 SD.

Hint
  1. Work out how many SD each result sits from its own mean.
  2. Level 1 is (3.48 − 3.20) ÷ 0.08 SD from its mean.
  3. Compare that distance with the 2 SD and 3 SD limits.

Review Multirule interpretation

Problem 2 of 3 · Hints available

A chloride level 2 control has a mean of 116 mmol/L and an SD of 1.5 mmol/L. Its last four runs:

Run1234
Result117.8 mmol/L118.0 mmol/L117.7 mmol/L118.1 mmol/L

Which rule, if any, is met?

2-2s needs results beyond 2 SD. The highest here is +1.4 SD.

Four results in a row are all beyond +1 SD. That meets 4-1s even though none goes beyond 2 SD.

Ignored a sequence because every control was inside 2 SD

Four consecutive results beyond 1 SD on the same side of the mean violate the 4-1s rule even when none exceeds 2 SD, as with four sodium controls between +1.33 and +1.67 SD. Waiting for a 1-2s warning misses this small systematic shift.

No result goes beyond 2 SD, so there is no warning. The sequence meets a rule without a warning.

The results are +1.2, +1.33, +1.13 and +1.4 SD: four in a row beyond the same +1 SD limit. That meets 4-1s and indicates a small systematic shift.

Hint
  1. Convert each result to SD from the mean.
  2. Every result is between +1.1 and +1.4 SD.
  3. Look for a rule that counts results beyond 1 SD.

Review Multirule interpretation

Problem 3 of 3 · On your own

A creatinine level 2 control has a mean of 5.20 mg/dL and an SD of 0.18 mg/dL. Level 1 has been near its mean in both runs.

RunLevel 2 result
Yesterday, 15:005.60 mg/dL
Today, 07:005.63 mg/dL

Which rule is met?

Both results are beyond +2 SD, at +2.2 and +2.4 SD, in two runs in a row. That meets 2-2s, which does not require a result beyond 3 SD.

Looked only for a single value beyond 3 SD

Two consecutive results beyond 2 SD on the same side of the mean, across two levels in one run or for one level across two runs, violate the 2-2s rule even when neither exceeds 3 SD. Waiting for a 1-3s violation lets a systematic error go undetected.

(5.60 − 5.20) ÷ 0.18 = +2.2 SD and (5.63 − 5.20) ÷ 0.18 = +2.4 SD. One level above +2 SD in two consecutive runs meets 2-2s and indicates systematic error.

Neither result goes beyond 3 SD. The highest is +2.4 SD.

Review Multirule interpretation

Use it

Optional for this step.

For reference