From Wiki Journal Club
Jump to navigation Jump to search
Litton E, et al. "Intravenous iron or placebo for anaemia in intensive care: the IRONMAN multicentre randomized blinded trial". Intensive Care Med. 2016-09-30. 42:1715-1722.
PubMedFull textPDF

Clinical Question

In critically ill patients with anemia, does early intravenous iron compared to placebo reduce the need for blood transfusion?

Bottom Line

Among critically ill patients with anemia, intravenous iron did not decrease the amount of blood transfused.

Major Points

The safety of blood transfusion has markedly improved in recent decades, but still carries some risk. Use of supplemental iron to utilize the patient's own internal processes to replenish blood in times of anemia is an attractive option.

This randomized, placebo-controlled trial conducted in four ICUs in Australia explored if the supplementation with 500 mg of intravenous ferric carboxymaltose as compared to placebo, would decrease the number of units of packed red blood cells were transfused in each group. Intravenous iron did not significantly decrease the median number of units transfused per patient, with roughly half of each cohort receiving transfusions. The ICU mortality and length of stay were also similar between groups. There was a difference in hemoglobin at hospital discharge (107 g/L vs. 100 g/L), which reached statistical significance, but is not likely to be clinically relevant.

This trial is unfortunately likely underpowered and therefore inconclusive. The mean number of units transfused was less than half of the number used in the power calculation. Even though it did not find a statistically significant difference, the point estimates favoured the use of intravenous iron supplementation and further study is required to inform is iron is a safe choice.


As of May 2024, no guidelines have been published that reflect the results of this trial.


  • Multicenter, double-blind, randomized, placebo-controlled trial
  • N=140 critically ill patients with anemia
    • Intravenous iron (n=70)
    • Placebo (n=70)
  • Setting: ICUs in Western Australia
  • Enrollment: 2013-2015
  • Mean follow-up: All patients followed until discharge
  • Analysis: intention-to-treat
  • Primary Outcome: Reduction in RBC units transfused


Inclusion Criteria

  • Age ≥18 years
  • Admission within prior 48 hours
  • Anticipated to require critical care beyond the next day
  • Hgb <100 g/L within the prior 24 hours

Exclusion Criteria

  • Suspected or confirmed sepsis
  • Ferritin >1200 ng/mL
  • Transferrin saturation >50%

Baseline Characteristics

IV Iron Group shown, groups were similar

  • Demographics: mean age 58 years, 37% female
  • Physiologic parameters: mean APACHE II Score 12.2, mean SOFA 6.1
  • Admission source: Emergency Department 20%, Ward 7%, Operating theatre 71%, other hospital 1%
  • ICU type: General Surgical 13%, Cardiothoracic 43%, Trauma 29%, Neurosurgical 3%, Medical 13%
  • Anthropomorphics: Weight
  • Labs: mean Hemoglobin 89 g/L, Ferritin 317 ng/mL, Transferritin saturation 13%, C-reactive protein 111 mg/L
  • Interventions: Mechanical ventilation 64%, Vasoactive infusion 73%, Renal replacement therapy 4%


  • 500 mg ferric carboxymaltose in 0.9% saline as two consecutive 50 mL syringes
  • Placebo
  • Interventions repeated every 4 days while admitted to ICU, to a maximum of 4 doses


Comparisons are IV iron vs. placebo.

Primary Outcomes

Total RBC units transfused
97 vs. 136
Median RBC units transfused (IQR)
1 (0-2) vs. 1 (0-3) (RR 0.71, 95% CI 0.43-1.18; P=0.53)

Secondary Outcomes

Proportion of patients receiving transfusion
54% vs. 56% (RR 0.97, 95% CI 0.72-1.31; P=0.87)
median Hemoglobin at hospital discharge
107 g/L vs. 100 g/L (Difference 7, 95% CI 1.89-12.11; P=0.02)
ICU length of stay
6 days vs. 6 days (Difference 0, 95% CI -1.07to 1.07; P=0.70)
Hospital length of stay
15 days vs. 18 days (Difference -3, 95% CI -7.95 to 1.95; P=0.75)

Adverse Events

ICU Mortality
7% vs. 4% (RR 1.67, 95% CI 0.41-6.71; P=0.47)
Hospital Mortality
1-% vs. 9% (RR 1.17, 95% CI 0.41-3.30; P=0.77)
Nosocomial Infection
28.6% vs. 22.9% (RR 1.25, 95% CI 0.71–2.21; P=0.44)
Clinically confirmed Venous Thromboembolism
6% vs. 6% (RR 1.0, 95% CI 0.26–3.84; P=1.0)


  • Medical patients made up only small portion of the population (14%)
  • Possibly underpowered: mean units transfused (1.9) was less than half of the estimate used in the power calculation (4.0)
  • Primary outcome was not normal and analysis had to be changed pot-hoc
  • Iron product may not be available world-wide and full class-effect may not be expected
  • no hemoglobin threshold was set trigger transfusion and could have led to bias


  • Not clearly reported

Further Reading