Every letter we have printed

The gallbladder section says some of the excess is attributable to weight loss rather than the drug. If the drug causes the weight loss, is that not a distinction without a difference for the person who ends up in theatre?

The placebo arm rates are substantial and almost never quoted, which makes the active arm figures look more dramatic than the comparison supports. The difference between the arms is the finding; the raw active figure is the headline.

A design nobody has run: withdraw, follow, then offer resumption and follow again. It would answer the question most readers actually have, and it is longer and more expensive than anybody has been willing to fund.

Reporting the median rather than the mean would change the appearance of most of these results, because the distributions are skewed. Where both are available the gap between them is itself informative and it is almost never printed.

Your piece describes tapering as pharmacologically pointless and then spends three paragraphs making a case for it. Pick one.

On reference standards, the honest position for most of this market is that there is no pharmacopoeial monograph for the sequence in question, so the standard is a house lot characterised against itself. That is not a scandal, but it should be stated, because a purity figure traceable to a house standard is not traceable in the sense the word usually carries.

System suitability is the section I would put first rather than fifth. If the tailing factor and the replicate precision are outside limits, the purity figure below them is arithmetic performed on an unfit separation. Everything else in the report is downstream of that box, and it is the box most often omitted.

Your submission design has a hole in it. Eight vials from one lot cannot separate variation between laboratories from variation between vials, because you have no replicate within a laboratory to estimate the second. Two vials each is a start and it is not enough, and the honest conclusion from your table is that the four figures differ, not that the laboratories do.

The threshold discussion needs one more sentence. A reporting threshold and an integration threshold are different settings, and software will let you set them independently. A method can integrate down to 0.01 per cent and report nothing below 0.10 per cent, and the resulting certificate is honest, complete and impossible to compare with anything.

Spontaneous reporting systems undercount by a factor nobody can pin down, and the databases are used in public argument as though they were registries. A note on what a voluntary reporting system can and cannot support would be a useful reference entry.

Labelling changes are a poor proxy for evidence strength, since they follow regulatory process rather than publication. A label updated in one jurisdiction and not another usually reflects timing, and it is read as disagreement about the science.

Cost is the maintenance variable that never appears in a trial protocol and dominates every real decision. An indefinite commitment priced monthly is a different proposition from a course of treatment, and the literature is written as though the two were the same shape.

Your piece treats maintenance as a single question. In the accounts I read there are at least three: maintaining a weight, maintaining a metabolic marker and maintaining a behaviour. They come apart in practice and are almost always discussed together.

Where a study offered continued support to both arms, the comparison is between two managed groups rather than between management and nothing. That is a better design and a narrower question, and coverage rarely notes which was run.

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.

On our storefront we replaced the badge with a plain line of text giving the molecule, the date and the report number. Conversion fell for a fortnight and recovered, and the questions we now receive are better ones. I mention it because the commercial argument for badges is weaker than sellers assume.

A defence of the practice, since you have printed several attacks. Before badges existed there was no public evidence of any testing at all. The current arrangement is a poor version of a good idea and it is still an improvement on what preceded it.

A quiet objection to the whole framing. Blind purchasing treats suppliers as adversaries, which is a reasonable design assumption and a poor description of most of the trade. The programmes that have learned most have been the ones willing to ask questions as well as run tests.

I have run a small blind programme for two years at my own expense. The finding that surprised me was how little variation there is between suppliers compared with the variation between lots from one supplier. That is a boring result and I suspect it is the true one.

Thank you for printing the letters that disagree with the desk at the same length as the ones that do not. It is a small editorial choice with a large effect on how the column reads.

Panel names imply organ specificity that the analytes do not have. Several of the markers in a liver panel are produced elsewhere and rise for reasons unconnected with that organ, which is why the panel is read as a pattern rather than item by item.

Lipase and amylase are poor screening tests in an asymptomatic person, and elevations without symptoms are common and usually of no consequence. Ordering them routinely generates a great deal of anxiety and very little information.

Fasting status changes several common analytes substantially, and self-arranged testing frequently records it wrongly or not at all. It is the single most common preventable source of a surprising result.

A baseline drawn before anything changes is worth more than any single later result, because it converts every subsequent value into a delta. The people who most regret not having one are the people looking at a first abnormal result with nothing to compare it against.

Intervals differ between laboratories because the assays differ, and results from two laboratories cannot be compared without the intervals beside them. People compare them constantly and the difference is attributed to their own biology.

Supply interruptions are the commonest cause of the situation your article describes, and they are not a clinical decision at all. It is worth connecting the titration question to the supply reporting elsewhere in the issue, because for many readers they are the same story.

A short interruption and a long one are not the same event and are discussed as though they were. Missing one weekly dose leaves substantial exposure in place; missing six weeks does not. The threshold is a matter of arithmetic and nobody prints it.

Your piece treats the four-week step as arithmetic, and I accept the arithmetic, but my prescriber moved me up every two weeks and I reached the top dose without difficulty. I do not think the schedule is as constraining as you suggest.

A plateau is a physiological outcome rather than a failure of dosing, and the reflex to escalate at that point is worth examining. Your piece hints at this and could say it outright: the curve flattening is what the curve does.

The mechanistic account is assembled from studies in different species, at different exposures, on different endpoints, and it is presented as one continuous story. Each join in that chain is an inference and none of them is flagged.

Tachyphylaxis in the emptying effect is a real observation and it is the mechanistic basis for symptoms easing at a stable dose. Stating it explicitly would help readers understand why the advice is to hold rather than to stop.

Management strategies circulate in this area with very little trial support, and the honest way to present them is as reported practice rather than as evidence. Your piece does that; a great deal of adjacent coverage does not.

Background incidence is the missing comparator in nearly every discussion of rare events. Some of the conditions named in this context occur at a measurable rate in an untreated population of similar age and composition, and the rate is rarely quoted alongside.

Printing the reasoning rather than only the conclusion is what makes this publication worth the subscription. Please keep doing it even when it makes the pieces longer.

Vitamin B12 status is imperfectly captured by a serum level, and the confirmatory markers are available and rarely ordered. Where the question genuinely matters, the serum value alone is the weakest form of the answer.

Supplementation before a test produces a value that reflects the supplement rather than the status, sometimes for days. Any series that includes both supplemented and unsupplemented draws is not a series.

Comparing results across laboratories is where readers come unstuck, because the interval belongs to the method and the method differs between analysers. A value that reads high against one laboratory’s range can sit comfortably within another’s, and neither laboratory is wrong.

A properly blind purchase means buying as an ordinary customer, from the ordinary catalogue, with no indication that the vial is going anywhere but a domestic address. Anything less and the supplier has had an opportunity, however small, to treat that order differently.

Reporting the negative results is what would distinguish a serious programme. Every honest series contains determinations that found nothing interesting, and a publication history composed entirely of findings is a publication history that has been curated.

You describe the services as competitors. From where I sit they are also each other’s only check, because a buyer who disagrees with one result has nowhere to go but another of the four. That is a peculiar market structure and it is worth naming as one.