Every letter we have printed

Your blind duplicate exercise is the most valuable thing this publication does, and eight vials is not enough to carry the conclusion you drew from it. Without replicates within each laboratory you cannot separate between-laboratory variation from vial-to-vial variation, and the honest reading is that four figures differed.

The word independent is used for two different things: independence of ownership and independence of the sample. A service can be entirely independent commercially and still be testing material the seller handed it, and readers hear the first and assume the second.

Absorptiometry divides tissue into three compartments and calls one of them lean, which includes water, organ and connective tissue as well as muscle. A fall in lean mass is not the same finding as a fall in muscle, and the reporting almost always translates one into the other.

The smallest change an instrument can resolve is the number readers most need and are least often given. Below that threshold a difference between two scans is not a small change; it is not a change at all, and treating it as one drives a great deal of unnecessary anxiety.

Where a trial protocol offered activity advice rather than a programme, the coverage reports it as a training intervention. Advice and supervision are different interventions with different effect sizes.

I have read your protein tables twice and I still cannot work out what I should eat. I appreciate that this is the honest position but it is not a useful one for a person in a supermarket.

A participant who is scanned is not a random participant. Substudy enrolment depends on site capability and on willingness, and the resulting sample can differ from the trial population in ways nobody records. It is a small caveat and it belongs in the methods paragraph.

The services could fix this unilaterally by serving badge images from their own infrastructure with the date rendered server-side. It is a fortnight of work, it costs a seller nothing, and it removes the possibility of a stale image entirely.

My objection is narrower than most of your correspondents’. I do not mind an undated badge on a page about a company. I mind an undated badge beside a price, because the proximity makes a claim about that product which nothing in the mark supports.

The lot number is the thread that holds a custody claim together, and it is frequently absent from the vial itself. Where the label carries only a molecule and a strength, there is nothing to connect the sample to any document at all.

On the archive point: a public record of submissions by vendor would be gamed within a month. Vendors would submit under the names of resellers, or through intermediaries, and the archive would show a distribution as selected as the current one but with a veneer of completeness.

Discontinuation for a good reason and discontinuation for a bad one are recorded identically in most datasets. Somebody who stopped because they had reached what they wanted and somebody who stopped because they could not tolerate it appear in the same bar of the same chart.

Reasons are ranked in most reporting and the ranking is a function of the category list. Change the list and the ranking changes, which means the headline finding of most of these surveys is a property of their questionnaire.

As a prescriber I would push back on your framing of the maintenance gap. We are not practising without evidence; we are practising on pharmacological inference, which is what clinicians do in every field where the trial has not been run. Calling it unevidenced makes reasonable practice sound reckless.

I would like to see more attention to what people were doing before, not only after. A maintenance phase entered from a long stable period looks different from one entered after years of oscillation, and baseline history is almost never reported.

The published trials in this class collected some of this data and reported very little of it, which is a genuine gap rather than a negative finding. It should be described as unreported rather than as unchanged.

Acute-phase effects move several micronutrient markers independently of status, so a value drawn during any inflammatory episode is difficult to interpret. Measuring an inflammatory marker alongside is standard practice in the literature and rare in self-arranged panels.

Frequency deserves a word. Repeating a panel monthly during a period of change produces a series in which most movement is variation rather than trend, and reading each report as news is a recipe for chasing noise.

You give the reference change value for ALT as about sixty per cent, which strikes me as so large as to make routine monitoring of it pointless. Is that your position?

Where a supplier does record temperature, ask for the file rather than the summary. A single maximum figure discards the duration, and duration above a threshold is what determines the loss.

Packing configuration should be documented and almost never is. The number and placement of the coolant packs determines how long a parcel holds temperature far more than the quantity of insulation, and it varies between consignments from the same supplier.

You draw a distinction between retest date and expiry date and then say suppliers use the wrong word. Which word do you think they should use, given that most of them have no study behind either?

The instrument-class section is the most useful thing you have printed on this subject. A unit-resolution quadrupole cannot distinguish a deamidation from the parent, and a great many identity confirmations in this trade are performed on exactly that class of instrument. The report says the mass matched, and it did, at the resolution available.

External calibration performed at the start of a batch will drift over a working day on some instruments. A laboratory quoting accuracy to a few parts per million on a sample run eleven hours after calibration is quoting a specification rather than a measurement, unless a lock mass was used.

Of the documents I have from companies in your dossier programme, the better ones state the observed mass alongside the theoretical value and name the instrument class. That is the standard the rest of the trade could meet tomorrow, and the fact that some suppliers already meet it is the useful part of your reporting.

Filtration through a sterilising-grade membrane is not sterilisation of the product unless everything downstream of the filter is controlled. The filter is one step in a system, and a filter certificate on its own supports a much narrower claim than the marketing usually built on it.

Media fill frequency and the number of units filled are the two figures that characterise an aseptic operation, and they are the two that never leave the building. A company confident in its suite could publish them tomorrow with no commercial exposure whatever.

Depyrogenation of glassware is a defined process with defined conditions, and a vial that has been washed rather than depyrogenated is a different container. Nobody in this market states which was done.

Water is the usual source. A synthesis and purification chain that uses water of unspecified quality has an endotoxin exposure at every step, and none of it is visible in a purity determination.

The recombinant reagent alternatives have matured and are accepted in several pharmacopoeias now. That development removes the last practical objection anybody raised about supply of the traditional reagent.

Escalating in response to a plateau is discussed as though the relationship between exposure and effect were linear beyond the studied range. It is not, in any published dose-response data I have seen, and the curve flattens there too.

The plateau discussion in community settings has a strong survivorship character. People whose trajectory continued do not post about a plateau, so the visible discussion overstates how universal the experience is.

The claim that nobody has randomised escalation intervals is too strong. There are protocol amendments in several programmes that effectively created slower-titration cohorts, and some of those have been analysed post hoc.

The supply-side reason for holding is never mentioned and is obviously real: a slower schedule costs less. In a market where people pay directly, economics is part of every schedule decision.

There is a version of this market with a certification tier, where a supplier pays for a fuller panel and is listed separately. It has not emerged, and it is worth asking why nobody has attempted it.

A note on scale. A supplier filling thousands of units has a fixed-cost argument for better process control that a small operation does not, and the market’s structure therefore affects what is even feasible. That is worth a piece in its own right.

Peptides frequently interfere with the amoebocyte lysate assay, and the interference can go in either direction. A laboratory that has not performed and reported the inhibition and enhancement study has produced a number of unknown validity, and that study is a routine part of the method.

The value I most want on an identity report is the theoretical mass the analyst calculated, printed beside the observed one. Half the reports I receive give only the observed figure, which requires the reader to redo the arithmetic and to guess which charge state and which counter-ion the analyst assumed.

A small format complaint. Reports that state a mass to four decimal places and an error in parts per million, without stating the mass at which the calibration was performed, are precise about the wrong thing. Precision presented without its provenance reads as authority and is not.

Your piece treats the source as a fixed component. It is a consumable with a maintenance history, and a fouled source changes response for some compounds far more than others. Ask when the source was last cleaned, and you will occasionally learn more than the spectrum tells you.