Mean arterial pressure in vasodilatory shock

Method

Mean arterial pressure, usually written MAP, by intra-arterial line or oscillometric monitoring, wherever RAISING it to a haemodynamic target is the therapeutic goal. That covers vasodilatory shock — distributive, septic, vasoplegic — and equally the vasopressor and vasoplegia-prophylaxis trials that share its sign: intraoperative vasoplegia, vasopressor-sparing adjuncts, and rescue for refractory hypotension in cardiac surgery or liver transplantation. THE PATIENT NEED NOT MEET A FORMAL SHOCK DEFINITION. The test is the direction of benefit, not the diagnosis. The timepoint (post-infusion, 1 h, 6 h) and the vasopressor background are recorded on the edge: the same rise means different things at the start of a noradrenaline infusion and at the end of a taper.

The method is part of the endpoint identity. Studies measuring the same quantity by a different method are held as separate endpoints.

Notes

SIGN INVERSION against CAT:outcome/mean-arterial-pressure, and this node exists because that one asked for it. Its notes say edges from vasopressor and vasodilatory-shock trials MUST NOT attach there, and that the endpoint should be split by clinical context before it carries many edges. Minted 2026-08-26 under ADR-0046 rule 4, which is additive — the sibling is the remedy and widening the parent is forbidden, because widening would silently re-label its existing edges as covering shock, which none of them do. Those existing four are hypertension trials, all no_effect. ADR-0014's floor was checked rather than assumed: PMIDs 38216185, 39615751 and 38904978 measure it in this batch, and the corpus already holds 36915146 (methylene blue in septic shock, RCT) and 37147207 (hydroxocobalamin for vasodilatory hypotension), so it is queryable across subjects rather than an artefact of one. THE SPAN CANNOT DISCRIMINATE THIS NODE FROM ITS PARENT AND THAT IS DELIBERATE. Both are called "mean arterial pressure" and both are abbreviated MAP; what separates them is the POPULATION, which no object span carries. "MAP" is written into this measurement on purpose so that the abbreviation ties between the two nodes instead of resolving only to the parent — spanNamesOutcome passes on a tie, and a tie is the honest verdict when the span is genuinely compatible with both. Do NOT "fix" this by declaring an abbreviation here: uniqueness would force it off the parent and break every hypertension span. The layer that actually discriminates is the refuter reading these notes, which is how both claims that motivated this node were caught. Contrast CAT:outcome/hospital-length-of-stay, where the discriminating token IS in the span ("ICU LOS" versus "hospital LOS") and an abbreviation of the shared substring was therefore removed. WIDENED ON ITS FIRST DAY, and the reason is recorded because it was a defect in the mint rather than a change of mind. As first written this node said "vasodilatory shock" and nothing else, while the parent's exclusion names "vasopressor AND vasodilatory-shock trials" — so it was narrower than the exclusion it exists to receive, and a trial of prophylaxis against intraoperative vasoplegia fell in the gap. PMID 37101089 is that trial: its primary outcome is noradrenaline dose needed to hold MAP above 65 mmHg, its patients are not in shock, and BOTH extraction passes filed it on the parent. Their reading was defensible against what this node said, which is what makes it the node's fault. Widening cost nothing because the node had zero edges — ADR-0046 rule 4 forbids widening a node to swallow a near-miss because that silently re-labels its existing edges, and there were none to re-label. The same widening applied to the PARENT would still be forbidden and always will be.