Reviews/Heart Failure Management/Intravenous iron reduces heart failure hospitalisation
Iron · HFM-IRON-01 · v2
Borderline: within 4 points of the Likely/Uncertain boundary, so a re-grade could land either side.
Several randomised trials tested whether giving iron into a vein lowers the number of times heart failure patients with low iron are admitted to hospital. One dedicated trial found fewer such admissions with iron [7], and another leaned in the same favourable direction and became statistically convincing only after an analysis that set aside the pandemic period [1]. The single largest trial, however, found no clear difference overall [9], and one trial's benefit was seen mainly in patients whose heart failure came from blocked arteries, not others [7]. A real-world record of patients given iron during admission could not test the question because it had no comparison group [8]. The panel judged the direction of the evidence consistently favourable but not firmly settled, because the two biggest trials did not meet their main targets and all patients studied had the reduced-pumping form of heart failure.
Three independent dedicated randomised trials point the same way on heart failure hospitalisations. AFFIRM-AHF reduced HF hospitalisations, with the effect driven by the ischaemic subgroup (RR 0.65, 95% CI 0.47-0.89, p=0.007) and absent in non-ischaemic patients (RR 1.11, 95% CI 0.75-1.66, p=0.60; pinteraction 0.039) [7]. IRONMAN's primary composite of recurrent HF admissions and cardiovascular death gave RR 0.82 (95% CI 0.66 to 1.02; p=0.070), reaching significance only in the prespecified COVID-censored analysis (RR 0.76, 95% CI 0.58 to 1.00; p=0.047) [1]. HEART-FID (N=3065) recorded 297 versus 332 HF hospitalisations but was neutral on its hierarchical composite (win ratio 1.10, 99% CI 0.99 to 1.23; threshold 0.01) [9]. The single-arm real-world cohort [8], with 63% rehospitalised within 12 months and no comparator, documents residual risk but cannot demonstrate a reduction. The contested point: The evidence divides on whether the favourable direction amounts to an established reduction: AFFIRM-AHF is positive specifically on HF hospitalisations [7], yet IRONMAN missed its primary endpoint at p=0.070 [1] and the largest trial, HEART-FID, was neutral [9]. It divides again on population, since every trial enrolled HFrEF only [1][7][9], and on whether AFFIRM-AHF's benefit generalises beyond its ischaemic subgroup [7].
All 9 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 Heart Failure Management, 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 dedicated hard-outcome trials missed significance: IRONMAN's primary endpoint was p=0.070 and reached significance only in a prespecified COVID-censored sensitivity analysis [1], and HEART-FID (N=3065) was neutral on its hierarchical composite [9].
AFFIRM-AHF's hospitalisation benefit was concentrated in the ischaemic subgroup (RR 0.65, 0.47-0.89) with a null non-ischaemic estimate (RR 1.11, 0.75-1.66, NS; pinteraction 0.039) [7], limiting generalisability of the reduction.
Where it remains uncertain. The largest trial was neutral (win ratio 1.10, 99% CI 0.99 to 1.23) [9] and AFFIRM-AHF's benefit was confined to the ischaemic subgroup (non-ischaemic RR 1.11, 95% CI 0.75-1.66; pinteraction 0.039) [7].
Aleatory
Primary estimates sit at the margin of significance (IRONMAN p=0.070; HEART-FID win ratio 1.10, 99% CI 0.99 to 1.23), so chance variation in event counts could move them either side of their thresholds [1][9].
The load-bearing papers from a corpus of 9, each classified by what it does to the claim. Every DOI is taken from the corpus the claim was graded on.
Largest ferric derisomaltose RCT; favourable direction on recurrent HF admissions + CV death, primary NS but COVID-sensitivity significant
“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). In the COVID-19 analysis, 210 primary endpoints (22·3 per 100 patient-years) occurred in the ferric derisomaltose group compared with 280 (29·3 per 100 patient-years) in the usual care group (RR 0·76 [95% CI 0·58 to 1·00]; p=0·047).”
Panel auditInternal validity: high · scope of inference: Patients with heart failure, LVEF ≤45% and iron deficiency (transferrin saturation <20% or ferritin <100 μg/L) in the UK.Key limitations: open-label design · primary endpoint not significant (p=0·070) · composite outcome combines HF admissions with CV death
Positive dedicated trial on the exact outcome; note benefit concentrated in ischaemic subgroup (RR 0.65, 0.47-0.89)
“In AFFIRM-AHF, intravenous ferric carboxymaltose (FCM) reduced heart failure (HF) hospitalisations and improved quality of life versus placebo in iron-deficient patients stabilised after an acute HF episode.”
Panel auditInternal validity: medium · scope of inference: Iron-deficient patients stabilised after acute HF, split by ischaemic versus non-ischaemic aetiology.Key limitations: subgroup analysis · effect confined to ischaemic patients (pinteraction=0.039) · composite combines HF hospitalisations with CV death
Largest trial (N=3065): numerically fewer HF hospitalisations with FCM but neutral on hierarchical composite (win ratio 1.10, 99% CI 0.99-1.23) — tempers the effect size
“a total of 297 and 332 hospitalizations for heart failure, respectively, occurred by month 12”
Panel auditInternal validity: high · scope of inference: Ambulatory patients with heart failure, LVEF ≤40% and iron deficiency receiving standard therapy.Key limitations: hierarchical composite rather than isolated hospitalisation endpoint · win ratio 99% CI crosses 1 · HF hospitalisation difference not separately tested
Real-world single-arm data: rehospitalisation remained common; no comparator so cannot evidence reduction
“among 119 discharged alive, 75 (63%) were rehospitalized within 12 months”
Panel auditInternal validity: low · scope of inference: Real-world HFrEF patients with iron deficiency given intravenous iron during hospitalisation at a single centre.Key limitations: no control group · retrospective single-centre · small sample · incomplete dosing and outpatient continuation
4 of 9 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.
Consistent favourable direction and one clearly positive dedicated trial support a Likely verdict, but two of three major RCTs missing their primary endpoints and HFrEF-only populations keep confidence at the lower 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.
HFpEF evidence absent. The claim is stated for heart failure broadly but every trial enrolled HFrEF (LVEF ≤40-45%) [1][7][9], so the preserved-ejection-fraction population is untested. The panel rates this the highest-importance open gap for the claim, and names the study design that would close it: A dedicated randomised trial of IV iron versus placebo in iron-deficient HFpEF patients with HF hospitalisation as a prespecified endpoint.
Open question: Does IV iron reduce HF hospitalisations in patients with preserved ejection fraction? Why it is still open: Every eligible trial enrolled HFrEF only (LVEF ≤40-45%) [1][7][9]; the corpus contains no HFpEF data.
Priors is a sceptical first reader, not peer review, and not medical advice.
Release 2026-q4 · graded 21 Oct 2026 · protocol v0.19