20, 19, 21, 23, 30
The dominant evidence base for ECPR patient selection and the backbone of Chapter 20 §20.5. Three practice-changing signals: signs of life are the strongest pre-cannulation variable and belong in the criteria list as a route in, not a footnote; the rhythm NOW carries more information than the rhythm the ambulance found; and transient ROSC roughly triples the odds of a favourable outcome and should be subtracted from elapsed low-flow time.
None — within-cohort comparisons among ECPR recipients only
UMIN000036490. Individual analyses carry their own DOIs — see Chapter 20 references 1–12
ECPR
Retrospective, registry, ECPR recipients only — there is no comparison group of patients who were refused, so every finding is prognostic and none is a treatment-effect estimate. Contributing units' own selection criteria are invisible in the data and self-confirming. Japanese practice may not generalise. All figures read from structured abstracts. The centre-volume analysis within this registry carries two numerical problems flagged in Chapter 20 §20.9 and not corrected.
SIGNS OF LIFE on arrival (n=1,395): present in 250 (17.9%); favourable neurological outcome 38.0% vs 8.1%, aOR 5.65 (3.97–8.03). FACTORS (n=1,823): shockable at scene OR 2.11 (1.16–3.95); shockable on arrival OR 2.59 (1.60–4.30); bystander CPR OR 1.63 (1.03–1.88); body movement OR 7.10 (1.79–32.90); gasping OR 4.33 (2.57–7.28); pupillary reflex on arrival OR 2.93 (1.73–4.95). TRANSIENT ROSC (n=1,501): 328 (22%); favourable neuro 26% vs 12%, aOR 3.34 (2.35–4.73); survival 46% vs 24%, aOR 3.99 (2.95–5.40). ARRIVAL RHYTHM in initial-shockable patients (n=1,114): sustained shockable 22.4%, PEA 9.5% (aOR 0.35), ASYSTOLE 2.7% (aOR 0.08). CONVERSION to PEA (n=718): 12.9% vs 26.4%, aOR 0.42 (0.27–0.66); worse the earlier it happened. INITIAL ASYSTOLE (n=202): 12 (5.9%) favourable, 7 of them hypothermic; among non-hypothermic patients in asystole or PEA on arrival, ALL 107 lacking one or more of witness, bystander CPR, or signs of life/pupil under 5 mm had unfavourable outcomes. INITIAL PEA: favourable 8.2%; ACS 9.7%, PE 19.3%, NO SURVIVORS with acute aortic disease or primary cerebral disease. PROLONGED LOW-FLOW 60 min or more (n=708 non-hypothermic): 10.0% favourable. PUPILS (n=723 with negative light reflex and mydriasis on arrival): 10.2% still favourable; positive light reflex at ICU admission OR 11.3 (5.17–24.7). SEX (n=1,819): female aOR 1.60 (1.05–2.43). START-ECPR score (n=648 non-shockable): survival 4.4% / 10.7% / 30.8% at scores 0 / 1 / 2–3.
2,157 adults who received ECPR across 36 Japanese institutions between 1 January 2013 and 31 December 2018. Individual secondary analyses use subsets of 202 to 1,823 patients depending on the question.