Anker 2009.

Methods	RCT; parallel‐group (phase III)
Participants	Patients with chronic heart failure as determined by NYHA status or ejection fraction with iron deficiency determined by ferritin < 100 mcg/L or 100‐299 mcg/L if TSAT < 20% . Patients randomised at 75 sites in 11 countries
Interventions	Participants randomised to receive intervention (ferric carboxymaltose, 200 mg) or equivalent volume of placebo (0.9% sodium chloride) in a 2:1 ratio. Dosing was weekly until theoretical repletion achieved (week 8 or week 12), then monthly until end of follow‐up (week 24).
Outcomes	Primary end points were self‐reported patient global assessment and NYHA functional class. Secondary end points included 6MWT distance and HRQoL. We were unable to extract these data due to isolated graphical representation. We were able to extract ferritin, Hb concentration and TSAT at the end of follow‐up and these are included in the review.
Study funding arrangements	Study was sponsored by Vifor Pharma
Author conflicts of interest	Multiple study authors report speaking and advisory fees from multiple pharmaceutical companies, including the study sponsor. Multiple study authors report being employees of the company sponsoring the study.
Sample size	459 participants were randomised 2:1 to receive the intervention. Of these, a non‐anaemic subgroup that received the intervention (n = 146) or the control (n = 76) treatment was examined.
Notes	Study was registered in a public clinical trials registry (NCT 00520780)
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Unclear risk	Quote: "Using a central interactive voice‐response system, we randomly assigned eligible patients, in a 2:1 ratio, to receive either ferric carboxymaltose (provided by Vifor Pharma) or placebo (normal saline)" Comment: whilst a central interactive voice response system was used, no comment is made on how the sequence was actually determined.
Allocation concealment (selection bias)	Low risk	Central interactive voice‐response system used. Site‐specific investigators highly unlikely to have been able to influence this.
Blinding of participants and personnel (performance bias) All outcomes	Low risk	Quote: "Because ferric carboxymaltose is a dark‐brown solution that is easily distinguishable from the saline placebo, study personnel responsible for the preparation and administration of the study drug (including at least one physician) were aware of the group assignments and therefore were not involved in any study assessments. To ensure that patients were unaware of the study drug they were receiving, black syringes were used to administer the study treatment and a curtain (or something similar) was used to shield the injection site from the patient’s view."
Blinding of outcome assessment (detection bias) All outcomes	Low risk	Quote: "Participants were blinded using opaque syringe and a curtain to conceal the injection site. Hence, self‐reported outcomes can be considered blinded. For other outcomes, outcome assessors were not involved in study drug preparation or administration."
Incomplete outcome data (attrition bias) All outcomes	Unclear risk	Comment: whilst there was dropout for the study due to participant withdrawal and death, they performed an ITT analysis. Data for key outcome metrics (HRQoL, NYHA classification, 6MWT) only represented graphically, and we could not extract them.
Selective reporting (reporting bias)	High risk	Comment: whilst the study included a specific 'non‐anaemic' subgroup, data for all outcomes in this group were not reported for this manuscript. This includes EQ‐5D‐5L, self‐reported global assessment, NYHA functional class and 6MWT. Whilst these data have been reported in designated substudies for non‐anaemic patients, this is not amenable to extraction (Comin‐Colet 2013Filippatos 2013; Van Craenenbroeck 2013).
Other bias	Unclear risk	Comment: study authors published protocol ahead of time, and the study was listed on a clinical trials registry. However, the study was drug‐company sponsored, and no reference was made to the conditions of funding.