Severe disability deaths are 24 years earlier and 40% avoidable

The study concerns a gap of twenty-four years. What it concerns, more specifically, is a woman named - let us say - Grace, thirty-one years old, who has had a persistent cough for six weeks and a mother who has taken her to the surgery three times and been told, three times, that this is “just how Grace is.” The distance between the study’s headline figure and that surgery waiting room is the distance this analysis aims to close.

Twenty-four years is not a statistic that arrives from nowhere. It arrives from a thousand small refusals to look properly - refusals the researchers behind this study, reported in the Guardian, have now totted up and found that four in every ten of the resulting deaths need not have happened at all. That is the number that should stop a reader mid-sentence: not the twenty-four years, which sounds like fate, but the forty per cent, which sounds like negligence with its books kept in order.

Observe what happens, mechanically, when Grace coughs. A person without her disability, coughing for six weeks, is heard as a body reporting a fault. Grace, coughing for six weeks, is heard as a person with intellectual disability behaving as such people do - the cough absorbed into the diagnosis she already carries, rather than treated as new information. This is not cruelty of the dramatic kind; no one has decided Grace matters less. It is cruelty of the administrative kind, in which a symptom is filed under an existing category instead of opened as a new file. The doctor is not lying. The doctor has simply stopped seeing.

Compare Grace with her flatmate, if she has one without a disability, who reports the same cough. The flatmate is referred, scanned, treated, and the matter closes within a fortnight. Grace’s cough becomes, six months later, the case history a coroner reads. The two women have had, materially, the same symptom. They have not had the same disease, because the disease each one gets is manufactured partly by the body and partly by who is listening to it. This is the principle the study makes visible, and it is worth stating plainly: an avoidable death is not a death that medicine could not have prevented, but one that medicine was not asked to prevent, because the asking depends on someone believing the patient is asking something real.

Now add the second figure the study gives us, which sharpens the first rather than merely repeating it: people with severe intellectual disabilities from minority ethnic backgrounds die fourteen years younger than their white counterparts within the same disabled population. This is the illustration compounded. If Grace is heard imperfectly because she is disabled, a second Grace - call her Amara - is heard through two filters at once, disability and difference each licensing the professional to file the complaint under something already known rather than something needing investigation. The interpreter who is not booked, the family history not asked after, the assumption that distress in this household simply looks louder than it does in that one - each is a small mechanism, and none appears in any single doctor’s misconduct file, because no single doctor did anything a tribunal would recognise as wrong.

The objection worth taking seriously is that severe intellectual disability carries genuine, independent medical vulnerability, and that some of this gap is simply the arithmetic of frailer bodies. This is true and it is also beside the point, because the researchers have already separated the frailty from the failure: the forty per cent is the portion the study’s authors judge to have been preventable by ordinary good practice, not by heroic intervention. The frailty explains a shortened life. It does not explain a shortened life cut short again by a missed scan.

Translate the whole of it into the register of an ordinary week: somewhere this month a carer is trying, for the fourth time, to get a receptionist to understand that “this is new” is a different sentence from “this is Grace.” Whether that sentence is heard is not decided by policy documents. It is decided in the sixty seconds after it is spoken, by whoever happens to be listening, and by whether the system has trained them to listen at all.