An approved medicine is only as good as the batch that reaches the patient

The argument

Proponents of this position start from a distinction between what a licence evaluates and what a patient swallows. Approval rests on a dossier and on a small number of batches made under close attention, often in a development facility, by people who know they are being watched. The medicine that reaches a patient years later is made continuously, at scale, in plants that may have changed hands, retooled or taken on other products since. Nothing in the original decision governs that later batch. On this reading a licence certifies a formulation, and the quality of the thing actually dispensed is a separate question that has to be answered again every time it is made. The argument is that the failures available at this stage are both common and severe. A sterile injectable can be contaminated. A tablet can be made with too much or too little active ingredient, or with the right amount unevenly distributed across a batch. Stored or shipped badly, a stable compound degrades into something else. An impurity introduced by a change of solvent can carry its own toxicity, undetected because nobody was looking for a substance the dossier never mentioned. Proponents emphasise that none of these are failures of the science behind the medicine; they are failures of execution, and they are invisible to any process that examines the medicine in principle rather than the plant in practice. Holders of the view conclude that inspection is the only instrument that reaches this. Quality cannot be approved into existence or inferred from paperwork submitted by the party being assessed; it has to be looked for, on site, repeatedly, by someone with the authority to stop production. They add that this is also what makes the rest of the system legible, since a surveillance signal or a trial result means something only if the product taken was the product described.

Premises

Counter-arguments

Critics do not dispute that manufacturing matters and deny that it is the right organising priority. Inspection, they argue, is a quality assurance function that can be delegated to standards bodies and to the manufacturer's own audited systems, which is largely how it already works at scale, and a regulator that puts its scarce expertise there is spending it on the part of the problem that is best understood. They add that an impeccably made medicine that does not work, or whose harms were never characterised, has passed every test this position sets, which suggests the priority is misplaced rather than the concern mistaken.

Rejecting the premises

[Rejecting P2] Several of the listed failures, such as novel impurities and degradation, are identified by analytical science developed during development rather than by execution alone, so the distinction the premise draws is softer than it needs to be. [Rejecting P3] Audited quality systems, batch release testing and supply chain documentation detect many manufacturing failures without an inspector on site. [Rejecting C] Continuing inspection may be necessary without being the organising priority, since a correctly made medicine can still be ineffective or inadequately characterised.