Last updated
Levey-Jennings and Westgard in myLabCompliance vs. Peer Comparison
Levey-Jennings charts and Westgard rules judge your control results against your own lab's targets, run by run. A peer comparison program, such as Bio-Rad's Unity Interlaboratory Program or Unity Next Peer QC, judges your average against other labs running the same control lot. The first catches errors as they happen; the second reveals bias you cannot see from inside your own data. myLabCompliance provides the first, with Levey-Jennings charts and Westgard rule violation detection; its public pages do not describe a peer comparison program, so labs that want peer data keep one alongside it.
Independent comparison by myLabCompliance. Not affiliated with Bio-Rad. Bio-Rad capabilities are described from Bio-Rad's public documentation.
Two questions, two tools
Internal QC asks: is my test system behaving the way it usually does? Peer comparison asks: is the way it usually behaves correct?
A method can be perfectly stable and still read consistently high. Levey-Jennings charts built on that method's own mean will look excellent, because the targets absorbed the bias when they were set. Only an external reference, whether peer data, PT or a reference material, shows the offset. The reverse also holds: peer data reported monthly will not catch a reagent pack that fails at 7 a.m. today. Each tool covers the other's blind spot, which is why many labs use both rather than choosing one.
What CLIA expects from internal QC
42 CFR 493.1256(c) says control procedures must detect immediate errors from test system failure, environment and operator performance, and must monitor accuracy and precision over time. Paragraph (d)(10) requires a mean and SD, or equivalent statistical parameters, for each lot of quantitative control material. Paragraph (f) requires acceptable QC before reporting.
Notice that CLIA mentions accuracy as well as precision. Internal QC with lab-set targets mainly monitors precision and change. Accuracy over time is where external data, PT above all, plays its part. CLIA does not require peer QC comparison; many labs use it because it fills the accuracy gap between PT events.
What myLabCompliance provides for internal QC
From the myLabCompliance QC page:
- Levey-Jennings charts with multiple QC levels, target and control limit display, level selection, date range selection and chart legends.
- Westgard rule violation detection as part of the QC module.
- QC scheduling and control lot management, with package insert records holding levels, target means, high and low limits, units, lot numbers and expiration.
- QC failure investigation tracking with root cause, corrective action and resolution fields.
- The option to exclude a QC run from Levey-Jennings reports while keeping its QC log row.
- CSV export and PDF-oriented reports.
- IQCP workflows.
These cover the run-by-run part of the QC program and connect it to the rest of the lab's compliance records.
What a Bio-Rad peer program adds
According to Bio-Rad's Unity Interlaboratory Program brochure, labs submit QC results from Bio-Rad controls and receive reports on analytical performance compared with peers, posted on QCNet, plus on-demand InstantQC reports for troubleshooting. The InstantQC help page lists three statistics: SDI (bias versus the consensus group), bias (distance from target) and TEB% (total error against allowable limits based on biological variation). Unity Next Peer QC is described on its product page as offering instant peer reports with no submission deadline.
Bio-Rad also offers Westgard Advisor, which its help page describes as an optional feature of Unity Real Time online that recommends SPC rules and controls per run, using the lab's data, a selected TEa and a consensus group that can be a peer, method or all-labs group.
The key ingredient is the shared control lot. Peer comparison works because many labs run the same Bio-Rad lot, so differences between labs reflect their test systems, not the material.
Reading a Levey-Jennings chart well
A Levey-Jennings chart is only as useful as the review behind it. Several habits make the difference:
- Look at all levels together. A shift on one level only points toward a different cause than a shift on every level at once.
- Look across runs, not just at today's point. Several points drifting to one side of the mean deserve attention even when none breaks a limit.
- Mark events on the timeline: reagent lot changes, calibrations, maintenance, new control lots. Many shifts line up with one of them.
- Do not quietly delete points. If a run is excluded from statistics, record why. myLabCompliance lets a run be excluded from Levey-Jennings reports while its QC log row stays, which keeps the history intact.
- Recheck the targets after a control lot change. Limits carried over from the old lot can hide or exaggerate problems.
Westgard rules formalize these habits. Rejection rules such as 1-3s and R-4s mainly catch random error; rules such as 2-2s, 4-1s and 10x mainly catch systematic error. Which rules a lab applies should follow from the quality required for each test and the method's performance, and the choice should be written into the QC procedure.
Side by side
| Question | Internal LJ + Westgard (e.g., myLabCompliance) | Peer comparison (e.g., Unity Interlab / Unity Next) |
|---|---|---|
| Did this run fail? | Yes, immediately | Not designed for run release |
| Is there a shift or trend vs. my usual performance? | Yes, with trend rules | Visible over the report period |
| Is my method biased vs. similar labs? | No, targets are your own | Yes, via SDI and bias |
| Is my imprecision typical? | Shows your CV only | Compares your CV with peers |
| Does it need a specific control brand? | No | Peer group depends on using the program's control lots |
| Where do corrective action, documents and competency live? | In the same platform | Separate system |
Capabilities in this table are based on each vendor's public pages; confirm details with each vendor.
Do you need peer comparison?
It depends on the lab. Questions that help decide:
- How often does PT give you feedback on each analyte, and are there analytes PT does not cover?
- Do you run third-party controls with a large peer group, or manufacturer controls with none?
- Have you had PT failures that, in hindsight, were slow biases?
- Do you need a consensus group to set quality goals or support IQCP risk assessments?
- Would peer data help with the twice-yearly instrument comparison under 42 CFR 493.1281?
If most answers point to yes, peer data is worth keeping. myLabCompliance does not replace it; it handles the rest of the QC program and the surrounding records.
When peer data and internal QC disagree
Sometimes internal QC looks stable while peer reports show a growing bias, or the reverse. Treat the disagreement as information. Check that the right lot and peer group are selected, that units and method codes match, and that the data submitted is complete. If the bias is real, look at calibration, reagent lots and instrument maintenance, compare with recent PT results, and document what you found and did in a corrective action record. If no action is needed, record the reasoning and the reviewer. Over time, these notes build a record showing that the lab looks at external data and acts on it, which is exactly the kind of evidence inspectors ask about when they review how QC data is used.
Running both without double work
If you chart QC in myLabCompliance and also take part in a Bio-Rad peer program, data needs to reach Bio-Rad by whatever submission route Bio-Rad supports for your account. myLabCompliance's public pages do not describe sending data to Bio-Rad or importing from Unity; its QC module lists CSV export, so ask Bio-Rad what file formats it accepts for manual submission and test it before relying on it.
Alternatively, keep Unity as the daily QC record for analytes where peer data matters most and use myLabCompliance for other tests and for all the non-QC records. Whichever route you choose:
- Use the same lot numbers and targets in both places.
- Name the system of record for each test in the QC procedure.
- File peer reports in myLabCompliance document control with review status, so peer review is documented alongside QC.
- Note peer findings in QC investigations when they inform the cause.
Setting targets with both in view
Peer data is also useful when setting targets for a new lot. After a crossover, compare your new lab mean with the peer mean for the same lot. A large gap is worth explaining before the lot goes live. CLIA allows using an assayed control's stated value if it matches your method and instrument and you verify it (42 CFR 493.1256(d)(10)(ii)), but your own mean and SD usually give tighter, more useful limits. Enter the approved values in whichever system charts the QC.
Performance specifications verified under 42 CFR 493.1253 are the baseline: if peer data or PT shows the method has moved away from what you verified, that is a finding to investigate, not just a chart to adjust.
See myLabCompliance Levey-Jennings charts and Westgard detection on a sample of your QC setup.
FAQs
Does myLabCompliance include Levey-Jennings charts?
Yes. Its QC software page lists Levey-Jennings charts with multiple levels, target and control limit display and date range selection.
Does myLabCompliance detect Westgard rule violations?
Yes. Westgard rule violation detection is listed as part of its QC module.
Does myLabCompliance offer peer comparison?
Its public pages do not describe an interlaboratory peer comparison program. Labs that want peer data can keep a program such as Bio-Rad's alongside myLabCompliance.
Is peer comparison required by CLIA?
No. CLIA requires control procedures, acceptability criteria and PT for regulated analytes. Peer QC comparison is optional and helps detect bias between PT events.
What is SDI?
Bio-Rad's InstantQC help describes SDI as a measure of your bias compared with the selected consensus group.
Can I send myLabCompliance QC data to Bio-Rad?
myLabCompliance's public pages do not describe a Bio-Rad connection. Its QC module lists CSV export; check with Bio-Rad which submission formats it accepts.
Why do my LJ charts look perfect but PT is borderline?
Your targets may have absorbed a stable bias. Internal charts show change, not correctness; compare with peer data or PT to check accuracy.
Which Westgard rules should I use?
It depends on the quality required for each test and your method's observed performance. Document the chosen rules and reasoning in your QC procedure and review them when performance changes.
Sources
- eCFR 42 CFR 493.1256
- eCFR 42 CFR 493.1281
- eCFR 42 CFR 493.1253
- Bio-Rad, Unity Interlaboratory Program (PDF)
- QCNet help, InstantQC Reports
- Bio-Rad, Unity Next Peer QC
- QCNet help, Westgard Advisor online
- myLabCompliance, QC software
Related myLabCompliance pages
- Lab QC: Levey-Jennings and Westgard
- CLIA Levey-Jennings charts
- Westgard rules explained
- Best lab QC and compliance software guide
- Standalone QC management software for clinical labs
- Using myLabCompliance alongside Bio-Rad Unity
Bio-Rad, Unity, Unity Real Time, UnityWeb, Unity Next, QCNet, InstantQC, Westgard Advisor, EQAS, Lyphochek, Liquichek and InteliQ are trademarks of Bio-Rad Laboratories, Inc. myLabCompliance is independent software and is not affiliated with or endorsed by Bio-Rad. Bio-Rad product details here are taken from Bio-Rad's public pages and help documentation, linked in the sources; check them with Bio-Rad. This page is educational and is not legal or regulatory advice.