A laboratory accredited through the CAP Laboratory Accreditation Program is inspected against the CAP checklists rather than against the regulation directly. The checklists carry the CLIA requirements for introducing a new method, and then add to them. The CLIA requirements are the four or seven characteristics covered in what CLIA requires before you report a result.
The practical difference is not the length of the initial list. Several CAP requirements recur. The CLIA verification of performance specifications is finished when the method goes live. Analytical measurement range (AMR) verification repeats twice a year, per analyte, indefinitely. For a chemistry laboratory running a few dozen analytes, that is a continuing workload, and one that is easy to underestimate.
One caution before the detail. Checklist requirements are revised between editions, and they differ by discipline. Treat what follows as an orientation to what the studies are and how to analyse them. Then check the current checklist for your own discipline for the wording that applies to you.
The two terms are routinely used as though they were interchangeable. The confusion is understandable because the experimental procedure is often the same, but the two studies answer different questions.
Calibration verification asks whether the calibration still holds. Do materials of known concentration, run as patient samples, still recover their assigned values? The study is a check on the calibration itself.
AMR verification asks whether the method still recovers correctly across the whole analytical measurement range, the range it can measure directly without dilution or concentration. AMR verification is a check on the range rather than on the calibration. The materials must therefore include low, mid and high levels reaching towards each end of the range. See analytical measurement range and reportable range for how that range is set in the first place.
The distinction also shapes the analysis. Calibration verification compares each material with its assigned value. AMR verification asks whether recovery holds across the whole range, including its ends. A linearity study is the natural analysis where the materials are a dilution series with relative rather than assigned values. The deviation from linearity at each level is then what you compare against your allowable limit.
The CAP chemistry and toxicology checklist expects laboratories to verify the AMR at least every six months as a baseline. AMR verification is also expected when changes are made to an assay without recalibration.
There is a useful exemption. A separate AMR verification is not required where the laboratory already calibrates, or verifies calibration, at least every six months. The exemption holds only if the calibration material spans the low, mid and high points of the range. The calibration work has already answered the question. Laboratories that calibrate frequently can therefore avoid duplicating the study, provided the materials genuinely span the range. A calibration at two levels near the middle does not span it, whatever its frequency.
Requirements differ by discipline. Coagulation tests based on direct measurement of an analyte, those measuring activity or concentration by enzyme immunoassay, immunoturbidity or chromogenic methods, require AMR verification. Clot-based tests do not. Linearity studies are not required for the calibration and calibration verification of CBC instruments. Check the checklist for the discipline you are inspected under rather than generalising from chemistry.
The second recurring requirement catches laboratories with more than one analyser. Instruments measuring the same analyte must be checked against each other at least twice a year for comparability of results. The requirement comes from CLIA (42 CFR 493.1281), and the CAP checklists carry it. The requirement applies whether the analyte is measured on two instruments of the same model or by two different methods.
Instrument comparability is a different question from the method comparison you ran when the method was introduced. The acceptance criteria are different, and the analysis is easy to get wrong. Comparing instruments within one laboratory covers it in full.
As with CLIA, the checklist tells you the study is required and how often, not what result passes. The laboratory defines the acceptance criteria, and has to be able to justify them.
For AMR verification the criterion is the allowable deviation from the expected value at each level. That criterion should come from the same place as the rest of your performance specifications: an allowable total error figure, or a specification derived from biological variation. Deriving it from what the instrument happens to achieve is circular. A drifting method will pass a criterion set by its own drift.
The material has to behave like a patient sample in the method, or the recovery you measure is a property of the material rather than of the assay. Whether a material behaves this way is the commutability question, and the question applies as much to a linearity set as to a reference material. Routine quality control material is generally not suitable for verifying calibration unless the manufacturer specifically designates it for that purpose.
The levels should span the AMR rather than cluster in the middle, and should include points near each end, since the ends are where recovery fails first. Where the method is used near a clinical decision point, a level near that point is worth including even if it falls in the middle of the range.
Download the CLSI EP06-A linearity example workbook (.xlsx) — five IgM dilutions with the deviation from linearity at each one judged against a ±5% allowable limit, ready to open in the Analyse-it trial. Use this analysis when the AMR materials are a dilution series.
The example workbook is downloading.
It opens in Excel on its own — the data and the finished results are both in it. Analyse-it is what lets you change the analysis and re-run it, try the same study on your own data, or work through it to see how the software handles it.
Every feature from all five editions for 15 days.
Treating calibration verification and AMR verification as one requirement. They answer different questions. The exemption works in one direction only: spanning calibration can satisfy AMR verification, but a single-level calibration check does not.
Materials that do not span the range. Three levels clustered near the middle verify the middle. The ends of the interval are what the study exists to check.
Using QC material for calibration verification. Unless the manufacturer designates it for that purpose, routine control material is generally unsuitable, and a non-commutable material gives a recovery that says nothing about patient samples.
Setting the limit from the instrument’s own performance. The allowable deviation has to come from a clinical or biological requirement. A criterion derived from observed scatter will be met by definition.
Forgetting the second analyser. Comparability across instruments is its own twice-yearly requirement, and running AMR verification on each instrument separately does not answer it.
Analyse-it produces the analyses a CAP AMR verification and the twice-yearly instrument check rest on, inside Excel:
Every feature from all five editions for 15 days. Recovery, bias and linearity are in the Method Validation and Ultimate editions, from US$ 475 a year. Difference plots and Bland–Altman bias for comparing instruments are also in the Medical edition. Validated against NIST and CLSI reference datasets. For the requirements on introducing a method, which these add to, start at what CLIA requires before you report a result; for the twice-yearly check, see comparing instruments within one laboratory.