Reviews/Iron Status/Intravenous iron cuts heart-failure hospitalisation
Ivd · IRON-IVD-01 · v2
Borderline: within 4 points of the Likely/Uncertain boundary, so a re-grade could land either side.
Four large randomised trials tested giving iron into a vein to heart-failure patients who were low on iron, and every trial pointed the same way: fewer heart-failure hospitalisations in the iron group. One trial that counted heart-failure hospitalisations on their own found a clear reduction [3]. But the two biggest and most recent trials did not meet the main result they set out to prove [1][8], and another trial only became convincing after a planned re-analysis that removed the COVID period [2]. There are also signs that men benefited while women did not, and women even had numerically more deaths [4]. So the direction is consistent, but the strongest formal tests are not all lined up behind it, which is why the panel treated a reduction as likely rather than settled.
Across four dedicated randomised outcome trials [1][2][3][8] every heart-failure-hospitalisation estimate favours intravenous iron, with rate ratios or hazard ratios between 0.74 and 0.82. AFFIRM-AHF is the one result that isolates total heart-failure hospitalisations and reaches significance (RR 0.74, 95% CI 0.58-0.94, p=0.013) [3]. IRONMAN's primary composite was RR 0.82 (0.66-1.02, p=0.070), becoming RR 0.76 (0.58-1.00, p=0.047) only in the prespecified COVID-censored analysis [2]. FAIR-HF2's total heart-failure hospitalisations were RR 0.80 (0.60-1.06, p=.12) [1] and HEART-FID's hierarchical composite was null with an unmatched win ratio of 1.10 (99% CI 0.99-1.23) [8]. The contested point: The corpus divides on whether a uniformly favourable direction plus one clean significant heart-failure-hospitalisation result [3] outweighs two large recent trials missing their prespecified primaries [1][8]. It also divides on generalisability: the benefit is present in men (HR 0.74, 0.57-0.95) but absent in women (HR 1.07, 0.63-1.82) [4].
All 13 papers behind this claim with their DOIs and what each does to it, the panel’s arbitration, the evidence gap map and the open research questions. Free, and sent to your inbox.
We’ll also send you other claims from Iron Status, one at a time, and tell you when a release or a pivotal trial changes a grade. Unsubscribe by replying.
One reviewer's only job is to attack the finding. The challenges below were weighed before the grade was set.
The two largest and most recent trials fail their prespecified primary endpoints: FAIR-HF2 total HF hospitalisation RR 0.80 P=.12 [1] and HEART-FID hierarchical composite null [8]. A claim of reduction leans on directionally-favourable but non-significant primaries.
IRONMAN's primary was non-significant (p=0.070) and only crossed significance in a post-hoc COVID-censored sensitivity analysis (p=0.047) [2] — treating that as a positive result overstates the evidence.
Where it remains uncertain. The two largest and most recent trials missed their prespecified primaries — FAIR-HF2 total HF hospitalisations RR 0.80, p=.12 [1] and HEART-FID null with win ratio 1.10 (99% CI 0.99-1.23) [8] — while IRONMAN reached significance only after COVID censoring [2].
Aleatory
Several estimates sit against the null boundary — IRONMAN COVID RR upper bound 1.00 [2] and HEART-FID win ratio 99% CI 0.99-1.23 [8] — so play of chance cannot be excluded.
Epistemic
The two largest and most recent trials missed their prespecified primaries (FAIR-HF2 p=.12 [1]; HEART-FID null [8]), leaving the confirmatory question open.
Inferential
Most favourable endpoints are composites containing cardiovascular death; only AFFIRM-AHF isolates total heart-failure hospitalisation with significance [3].
The load-bearing papers from a corpus of 13, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Newest large RCT; consistent point estimate but non-significant and overall neutral, tempering the claim.
“The second primary outcome (total heart failure hospitalizations) occurred 264 times in the ferric carboxymaltose group vs 320 times in the placebo group (rate ratio, 0.80 [95% CI, 0.60-1.06]; P = .12).”
Panel auditInternal validity: high · scope of inference: Patients with HFrEF (LVEF ≤45%) and iron deficiency across 6 European countries.Key limitations: primary HF-hospitalisation endpoint not significant · composite first primary driven partly by CV death
Recurrent HF hospitalisation + CV death directionally reduced; primary non-significant, significant only in COVID-censored analysis.
“336 primary endpoints (22·4 per 100 patient-years) occurred in the ferric derisomaltose group and 411 (27·5 per 100 patient-years) occurred in the usual care group (rate ratio [RR] 0·82 [95% CI 0·66 to 1·02]; p=0·070).”
Panel auditInternal validity: medium · scope of inference: UK patients with HF, LVEF ≤45%, and ferritin <100 µg/L or TSAT <20%.Key limitations: primary endpoint not significant · open-label design · endpoint is a composite with CV death, not HF hospitalisation alone
Only trial isolating total HF hospitalisation as an endpoint; significant reduction directly supporting the claim.
“217 total heart failure hospitalisations occurred in the ferric carboxymaltose group and 294 occurred in the placebo group (RR 0·74; 95% CI 0·58-0·94, p=0·013).”
Panel auditInternal validity: high · scope of inference: Iron-deficient patients with LVEF <50% stabilised after an acute heart failure episode.Key limitations: primary composite endpoint p=0.059 (not significant) · HF-hospitalisation benefit is a secondary endpoint
3 of 13 papers shown. The full review lists all of them, with the panel’s arbitration →
Claims are graded at each quarterly release, and re-graded between releases only when a pivotal trial lands. Every change is a new version of the claim.
Uniformly favourable direction with one clean significant HF-hospitalisation result supports a likely reduction, but two recent large neutral primaries and sex heterogeneity keep it below the Established band.
Release 2026-q4 re-graded every claim from its corpus under protocol v0.19. An earlier release's grade is shown for the record; it came from an earlier protocol and is not a movement.
Assumptions that carry weight but are not themselves established.
The panel records the gap that would most move the claim, and the study that would close it.
No pooled synthesis of the four outcome trials. Each trial is individually underpowered on total heart-failure hospitalisation and their directions agree, but no meta-analysis in the corpus combines them into a confirmatory estimate. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: A prespecified individual-patient-data meta-analysis of FAIR-HF2 [1], IRONMAN [2], AFFIRM-AHF [3] and HEART-FID [8] on total heart-failure hospitalisation.
Open question: Does intravenous iron significantly reduce total heart-failure hospitalisation when the four outcome trials are pooled? Why it is still open: Direction is uniform (RR/HR 0.74-0.82) but only AFFIRM-AHF is individually significant on this endpoint [3], and FAIR-HF2 and HEART-FID missed their primaries [1][8].
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 18 Sep 2026 · protocol v0.19