Every letter we have printed

Nobody reports what maintenance costs in attention. The accounts I find most credible describe a low but continuous administrative load — ordering, storing, timing, monitoring — and that load is invisible in every clinical endpoint.

A randomised withdrawal design answers a narrower question than most readers take from it. It tells you what happens to people who reached a defined point and were then randomised, which is not the same as what happens to somebody who stops on their own terms.

I take this for kidney disease, not for weight. Every piece of writing I encounter about stopping is about the weight coming back. It has taken me a year to find anybody willing to say plainly that the renal benefit accrued over years of treatment and nobody has tested what happens if I stop.

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.

Cost is the leading reason people stop and it is discussed least, because it is not a clinical variable and therefore does not appear in clinical writing. Your treating it as a primary reason rather than a footnote is the right decision.

Reading these studies alongside your dossier work makes a point neither makes alone: the trial evidence describes a material of known provenance, and the market this publication covers does not guarantee that. Every extrapolation crosses that gap silently.

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.

A note on the buyers who do click through. Of the ones I have spoken to, most read the purity figure and stop. The method, the date and the sample identity are all on the page and are not being read, which suggests the reform needed is in what is emphasised rather than in what is disclosed.

A structural suggestion. If the services published the count of reports issued to each requester alongside the count published by that requester, the selective-publication problem would be visible without disclosing a single result. It is a number they already have.

The order of the questions matters. Whether training preserves composition during loss and whether it accelerates loss are different studies, and results from one are frequently used to answer the other.

Adherence to an activity component is measured badly in almost every study, usually by self-report, and the resulting variable is not comparable between trials even where both report it.

The control arm in a composition substudy is losing mass too, usually, and its composition ratio is the comparison that matters. Reporting only the active arm turns a comparative result into an absolute one.

One correction worth making loudly: the nominal mass on the label is not necessarily the mass in the vial. Where peptide content has not been determined, every concentration calculated from the label carries that uncertainty, and no amount of careful arithmetic removes it.

Displacement is small for these quantities and not always negligible. Adding 2 mL of diluent to a vial containing a few milligrams of solid gives very slightly more than 2 mL of solution. It is a fraction of a per cent and it is worth knowing that the direction of the error is always the same.

What I would fund, if I could fund anything, is a rolling blind purchase of the same molecule from the same source every quarter for three years. Longitudinal data on one line would settle more arguments than a hundred one-off determinations across the market.

Please repeat the exercise annually and publish the method conditions each laboratory used alongside the figures. The comparison becomes far more informative once a reader can see that the outlier ran a twelve-minute gradient and the others ran forty.

As a buyer I found the section on who chose the vial genuinely clarifying and slightly deflating. I have been treating vendor-published reports as equivalent to my own submissions for two years, and on your account they are not equivalent by an amount that cannot be measured.

Reconstitution before submission destroys the custody argument entirely, and buyers do it because a solution is easier to post than a vial. The convenience is real and it converts a question about a product into a question about a preparation.

An observation about scale. Custody discipline is easy at one sample a month and hard at fifty, which is why the programmes with the most data have the weakest records. The failure is a consequence of volume rather than of care.

The point your piece makes about documentation applies to custody with more force. A supplier that publishes its own handling record gives a buyer something to check; a supplier that publishes only a result gives them something to believe.

Aliquoting breaks the chain silently. If a buyer decants a portion into a second tube before sending it, the laboratory receives a sample the supplier never touched, and any dispute about the result is now unresolvable in both directions.

The price table is the most useful thing you have published this year and also the thing most likely to be quoted out of context by somebody selling a comparison service. You might consider a note.

Your suggestion that laboratories publish the fact of a submission while keeping the result confidential is the first proposal I have read in this area that a laboratory could actually implement without breaching a client obligation. I have raised it internally.

Placement carries meaning. A mark beside a price reads as a claim about that product; the same mark in a footer reads as a claim about the company. Sellers know this and choose accordingly, and no rule anywhere governs the choice.

The soft-tissue composition around the measurement site changes during the study, and that affects the density estimate through the reconstruction. It is a known artefact with a published magnitude and it is rarely adjusted for.

The skeletal findings in the trial substudies are modest and the follow-up is short, which is a fair summary and an unsatisfying one. Where the honest answer is that the duration studied is shorter than the duration of interest, saying so is better than filling the gap.

Composition ratios from trials that included a structured activity component are quoted in discussions where no such component exists. The trial measured a package and the figure is read as a property of one element of it.

Age and sex distributions in the substudies differ substantially from the populations the figures are quoted about, and composition change is strongly patterned by both. Anybody reading a single ratio across a general population is discarding the two largest sources of structure in the data.

Point-of-care gastric ultrasound is the assessment that actually answers the question, and availability rather than evidence is the limiting factor. Where it exists it replaces a rule of thumb with a look.

The guidance documents themselves are freely available and rarely read by the people arguing about them. Linking the primary sources, as your reference desk does, is worth more than another summary of them.

Distinguishing expected effects from something that needs attention is the most valuable service a piece like this can perform, and the useful discriminators are pattern and trajectory rather than severity at a moment. Anything that is worsening rather than settling is a different conversation.

Your figures show diarrhoea at thirty-two per cent and constipation at twenty-three per cent in the same trial arm. I assumed one of these was an error until your mechanism section. It would be worth putting that explanation before the table rather than after it.

Effects reported during titration and effects reported at steady dose are different populations of events, and pooling them across the whole study makes both harder to interpret.

Please repeat the exercise annually and publish the method conditions each laboratory used alongside the figures. The comparison becomes far more informative once a reader can see that the outlier ran a twelve-minute gradient and the others ran forty.

I would settle for a much weaker reform than blinding: a published statement of who selected the sample. It is one line, it is not contentious, and it would let a reader weigh every result on the page correctly.

Speaking as a former quality manager, the safeguard that matters is whether the analyst knows whose sample it is before the run. Blinding at the bench costs nothing and removes the whole class of problem your article describes. It is a question worth putting to each of the four.

Between-subject variability in exposure is large and rarely mentioned in summaries aimed at general readers. Two people on an identical dose can sit some way apart in plasma concentration, which is a better explanation of divergent experience than most of the ones in circulation.

Albumin binding as the mechanism behind the long half-life is worth expanding. The acylation is not incidental chemistry; it is the design, and it also explains why these molecules distribute the way they do. The pharmacokinetics were engineered before the clinical programme existed.

Mechanistic plausibility is used as evidence throughout the discussion of this class, and it is the weakest form of evidence available. Where an outcome study exists it should be cited instead, and where none exists that should be said.

Cost is the reason least often recorded and most often given in private. Every survey I have seen collapses it into an "other" category, and it is very likely the largest single entry in the list.