Severe bronchopulmonary dysplasia, dichotomised at 36 weeks

Method

Severe bronchopulmonary dysplasia at 36 weeks postmenstrual age as a BINARY endpoint, pooled as a risk ratio. THE GRADING CLASSIFICATION IS RECORDED ON THE EDGE AND IS NOT OPTIONAL: the NIH consensus definition and Jensen's respiratory-support definition disagree about which infants are severe, and a trial may also use its own. Where a paper states no classification, that is an unstated instrument rather than a near miss.

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

Notes

Minted 2026-08-27 under ADR-0046 rule 3. ADR-0014's two-paper floor was checked rather than assumed: PMID 40180500 reports severe BPD (Jensen grade 2 or 3 by mode of respiratory support) at adjusted RR 1.06, 95 percent CI 0.93 to 1.21, as the trial's primary outcome, and PMID 36943265 reports severe BPD at RR 1.17, 95 percent CI 0.97 to 1.41 across 3 trials, n = 1892. Both are DHA in very preterm infants, so like the vasopressor-discontinuation node this one is not yet demonstrated across subjects; if it never attracts a second subject that is a fact about which literature this corpus holds. IT SHOULD HAVE EXISTED ALREADY. CAT:outcome/death-or-severe-bpd-composite's notes stated that both of its components were recorded as their own nodes. Only one was. The composite was minted in the same round and nothing checks a claim a note makes about the rest of the ontology, so the assertion stood for nine days while the endpoint it named had no home. DISTINCT FROM CAT:outcome/bpd-severity-grade-ordinal, which is the same disease analysed as the ODDS OF A HIGHER GRADE across a four-level scale and reported as an odds ratio; this node is a dichotomisation of that scale and the two are not interconvertible. DISTINCT FROM the composite above, which adds death and exists because death is a competing risk. DISTINCT FROM any-grade incidence: a risk ratio for severe disease and a risk ratio for disease of any grade are different quantities in the same paper and can point different ways -- in PMID 40180500 severe BPD is 1.06 and not significant while any grade is 1.09 and is. THE LABEL SAYS "DICHOTOMISED" TO EARN ITS OWN NAME. Worded plainly as "Severe bronchopulmonary dysplasia at 36 weeks" the label failed the discrimination test: the ordinal node and the composite both carry severe, bronchopulmonary, dysplasia and 36 weeks postmenstrual age in their own scored text, so the label named three nodes and not itself. Rather than take a KNOWN_AMBIGUOUS override -- which the total-cholesterol, plasma-nitrate and mean-arterial-pressure nodes each needed -- the label states the property that actually separates it, which is that this is the dichotomy rather than the ordinal or the composite. A bare span reading "severe bronchopulmonary dysplasia" will still tie across all three and be decided by a reader; that asymmetry is the same one the MAP node documents and is the intended behaviour. THE SIGN TRAP ON THE ORDINAL NODE APPLIES HERE UNCHANGED. A risk ratio above 1 means MORE disease. Read the point estimate, never the prose: one meta-analysis in this literature describes as an inverse association a set of estimates that are all above 1.