No Method on the COA: What to Ask Before You Accept It

A certificate of analysis is a set of claims, and each claim has two parts: the result and the basis on which the result was obtained. When a line carries a number but not a method, only half the claim is present. The number may be entirely correct and the batch may be entirely acceptable. What is missing is the information that would let anyone else compare it with anything.

This is one of the most common omissions on certificates in this trade, and it is worth understanding why it happens before deciding what to do about it.

A result without a method is not comparable

Comparability is the property that makes a certificate useful over time. A buyer compares a new certificate against the previous one, against an incoming test result, and against the specification. All three comparisons assume the numbers were produced the same way. When the method is unstated, the assumption is unverified, and a difference of a few tenths of a point becomes impossible to interpret: it could be a real change in the material, or it could be two laboratories measuring different things with the same label.

The problem is sharper for content than for most lines. Purity measurements reported by the same technique in the same laboratory are usually comparable even without a stated method, because the technique is implicit in the way the result is expressed. Content determination is not like that. Two accepted methods for the same material can produce figures that differ by more than the difference between two consecutive batches, and both can be correct.

The two methods behind a UDCA content figure

For this material, two approaches are in ordinary use, and they measure different physical quantities. A volumetric determination measures the whole acidic function in the sample. A chromatographic determination measures the response of the main peak after separation. Where the sample contains material that is chemically similar but chromatographically resolved, the volumetric method counts it and the chromatographic method does not.

The consequence is that the two methods answer slightly different questions. The volumetric figure describes the total acidic content of what was weighed; the chromatographic figure describes the proportion of that material which behaves as the main component under the separation conditions. Both are legitimate; neither is wrong. But a buyer who compares a volumetric result against a chromatographic result will see a discrepancy that has nothing to do with the material. Assay methods that separate covers the selectivity requirement in full; the relevant point here is that the label on the number matters as much as the number.

What each omission costs you

Line	If the method is stated	If the method is missing
Content	Comparable with your own result and with previous batches	A discrepancy cannot be attributed to the material or to the method
Related substances	Limits can be matched to the separation actually used	You cannot tell which impurities the panel would have detected
Loss on drying	Drying conditions known, so the volatile figure is comparable	Two drying regimes give different results on the same material
Residual solvents	Technique and any headspace conditions are known	Volatile losses during preparation cannot be excluded
Identification	The comparison basis and standard are traceable	The test becomes a statement rather than a comparison

The right-hand column is why this matters even when the batch is fine. A certificate with unstated methods cannot be used to build a trend, cannot be used to resolve a dispute with an incoming laboratory, and cannot be defended in a supplier audit where the auditor asks how a limit was evaluated.

Why the omission is common, and usually not deliberate

In most cases the method column is empty because the certificate template was designed around the specification rather than around the test. The specification lists the tests and the limits; the method sits in the laboratory's working instruction, where the technician can see it, so nobody feels the need to repeat it on the report. The document ends up technically true and practically incomplete.

There is a second and less innocent cause, which is worth being aware of but should not be assumed. A method that has not been validated for the material is sometimes not named because naming it would invite questions about its qualification. The distinction between the two causes is settled by asking for the method reference, not by inferring motive. Laboratories that have nothing to hide answer the question in a day.

The five questions worth sending

When a certificate arrives with unstated methods, a short list of questions resolves it without turning into an audit.

First, which method produced the content figure, and where is it described. Second, which standards were used, and were they valid on the day of testing. Third, for the impurity panel, whether the stated limits were evaluated against reference spots or against a calibrated system. Fourth, for the volatile figure, the drying temperature, the time and the sample weight. Fifth, whether the laboratory that produced these figures is the same one that produced the previous batch's figures.

The fifth question is the one buyers omit and regret omitting. A change of laboratory explains a step change in results far more often than a change in the material does, and it is invisible on a certificate unless somebody asks.

What we put on the certificate, and why

Our certificates name the method or its reference against every test, state the drying conditions alongside the volatile figure, and identify the laboratory that performed each test. Where a result sits close to a limit, we attach the trace or the calculation rather than only the number. None of this changes the specification. It changes what a reviewer can do with the certificate without writing back to ask.

For buyers who work with more than one testing laboratory, the same principle runs through third-party testing: the report is only as useful as its scope statement, and a result whose method is unstated has a scope that cannot be checked. Kept alongside the nine lines that carry weight on a certificate, that habit settles most method questions before they become discrepancies.

Two laboratories, one material, two different figures

The situation that makes a missing method statement expensive arrives early in a supply relationship. A buyer's own laboratory reports a content figure, our certificate reports a different one, and the difference is a few tenths of a percentage point. Neither laboratory has made an error. Without knowing which method produced each figure, there is no way to establish that.

Suppose a volumetric determination returns 100.2 per cent and a chromatographic determination performed on the same material returns 99.8 per cent. The volumetric method measures the whole acidic function of what was weighed. The chromatographic method measures the response of the main peak after separation, which excludes material that is chemically similar but resolved by the column. A difference of this size between the two is ordinary, and it is not evidence that anything is wrong.

With the methods named, that comparison takes ten minutes and closes with a note in the file. Without them, the same comparison takes a week, produces two more determinations, and ends in a judgement about which laboratory is more trustworthy, which was never the question.

The lesson generalises beyond content. Every line on a certificate where more than one accepted technique exists carries the same exposure, and the lines where the exposure is largest are those where results are used for release decisions rather than for information.

What a method statement should contain

A complete method statement is short and specific. It names the technique, identifies the method by reference to a documented procedure, and states the key conditions where those conditions affect the result. For a chromatographic determination that means the column and the detection wavelength and whether a correction was applied. For a volumetric determination it means the titrant and the indicator or the instrumental end-point. For the volatile figure it means the temperature, the time and the sample weight.

What a method statement does not need to be is long. A reference to a laboratory procedure with the key parameters stated is enough for any reviewer to judge comparability. What it cannot be is absent, because an absent method statement converts a measurement into an assertion.

There is a related practice worth adopting, which is to state the laboratory that performed each test. Where a single organisation performs all the testing, that is one line. Where testing is distributed, naming the laboratories turns an unanswerable question about comparability into a documented fact. Third-party testing deals with the scope question that follows; the method statement deals with the comparability question that comes first.

What goes into an incoming specification

A specification written for incoming material is not the same document as the specification the material was released against. The incoming document has to be capable of producing a decision, which means it has to be explicit about method comparability and about what happens when the methods differ.

Three clauses carry most of that weight. The first states which method is expected for each test, or accepts either of two named methods and says so. The second states the action when results differ by more than a stated amount, which is usually a split-sample confirmation rather than a rejection. The third requires the method reference to appear on the certificate as a condition of acceptance, which is the only clause that makes the other two enforceable.

That third clause is the one buyers most often omit and the one that saves the most time. A certificate without method references is not defective under most regulatory frameworks, and the guidance documents that describe certificate expectations treat the method statement as part of the record rather than as a condition of the certificate itself. Making it a contractual condition is how a buyer converts an expectation into an enforceable requirement.

Where the material has a monograph, the monograph itself provides the default method and the clause becomes a statement that the method used should be the one the monograph describes or an equivalent. Where the material has no monograph — as is the case for parts of this family — the clause does more work, because there is no published default for either party to fall back on. Methods that separate the analyte from its related substances are the ones worth naming in either case.

The published monographs are the natural starting point for that clause. The United States Pharmacopeia and the European Pharmacopoeia each describe, for the materials they cover, the technique that the monograph expects, and a clause that names the monograph's technique as the expected one is short and unambiguous. Where the material falls outside both, the clause has to be written from the technical file, and it should say so rather than invent a default.

A final practice worth adopting is to record what happened when the clause was first exercised. The first time an incoming result and a certificate disagree, the correspondence that resolves it is the most useful page in the file, because it establishes how the two organisations will handle the same situation in future. Buyers who keep that page spend less time on the second occurrence than on the first.

One last observation about method statements, and it is the one that decides whether the practice survives contact with a busy laboratory. A method statement is only useful if it is consistent. A certificate that names the method for some tests and omits it for others creates the impression, however unfairly, that the omissions are deliberate. Consistency is cheap and the alternative is expensive to explain, and it is the reason we treat the method column as a release item rather than as a formatting preference.

Frequently asked questions

Is a missing method statement a reason to reject a batch?

Not by itself. It is a reason to ask a question before filing the certificate. The measurement may be sound; what is missing is the information that would let you compare it with your own result or with a previous batch.

Which methods can produce a UDCA content figure?

Two are in common use: a volumetric determination and a chromatographic determination. They measure the same quantity by different means and do not always agree to the decimal place, which is precisely why the method needs to be named.

Why do some laboratories omit the method column?

Usually because the certificate template was built around the specification rather than around the test. The method lives in the working instruction and the technician knows it, so the field never gets populated. It is an administrative habit rather than a deliberate omission.

How much can two acceptable methods differ?

Small but not negligible. Differences of a few tenths of a percentage point are ordinary between a volumetric and a chromatographic determination, and a difference of that size will look like a discrepancy if the two numbers are compared without knowing which method produced each.

What should a complete certificate state for each test?

At minimum the test name, the limit, the result and the method or its reference. Where a result sits close to a limit, or where the test is the basis of a decision, the method reference is the part that carries the weight.

Can I repeat the test myself instead of asking?

You can, and for a material you buy repeatedly that is often worth doing. But a repeat performed by a different method still leaves the original figure unexplained, so asking and repeating are complementary rather than alternatives.