Intravenous iron should not be administered before confirmation of iron deficiency. Ferritin of less than 30 mg/mL was typically used in the studies included in the evidence base to identify women who would benefit from intravenous iron. If ferritin testing is not available, a peripheral blood smear demonstrating microcytic hypochromic anaemia typical of iron deficiency, while other causes of anaemia are reasonably excluded, can be an alternative. Alternatively, health workers may consider a treatment test where an improvement of anaemia is demonstrated after 2 weeks of oral iron treatment, if that is feasible.

• Oral iron may not be tolerated because of side-effects, and may be contraindicated or not absorbed (e.g. pre-existing condition leading to poor intestinal absorption).
• Rapid correction of anaemia may be required for women with severe anaemia close to the time of birth to avoid adverse outcomes from postpartum haemorrhage.
• Women need to be monitored during intravenous iron infusion and for at least 30 minutes after infusion for anaphylactic reactions; health systems need to have monitoring systems in place to track safety. Some intravenous preparations (e.g. iron dextran) appear to increase the risk of anaphylactic reactions more than others and should be avoided. The risk of anaphylaxis is estimated to be 9.8 per 10 000 for iron dextran, 1.5 per 10 000 for ferric gluconate, 1.2 per 10 000 for iron sucrose and 0.8 per 10 000 for ferric carboxymaltose.
• Intravenous iron preparations have been approved for treatment of iron-deficiency anaemia but not specifically for iron-deficiency anaemia in pregnancy. Therefore, treatment with intravenous iron should be confined to the second and third trimesters if the benefit is judged to outweigh the potential risk for both mother and fetus. The decision to use intravenous iron therapy should be made with the informed consent of the woman, ensuring that she fully understands the potential benefits and risks.
• Studies mostly used intravenous iron sucrose or ferric carboxymaltose. Iron sucrose was usually given as several 200-mg infusions on alternating days until the calculated iron deficit was reached. Ferric carboxymaltose was usually given as single infusion of 20 mg/kg body weight or up to a maximum of 1000 mg.
• The use of intravenous iron increases costs (related to both medication and infrastructure), and so the frequency of its use will need to be considered in the context of resources available.