The review addressed clear research questions with clear inclusion criteria. Several sources were searched and attempts were made to obtain unpublished data. Only English-language publications were included, which introduced a risk of language bias. Attempts were made to minimise bias and error in the review process. Trial quality was assessed using described criteria and the trials were deemed fair. It was unclear which quality criteria the trials did not meet. The lack of heterogeneity between trials suggested that the synthesis was appropriate.
The authors' conclusions reflect the evidence presented, nut results were based on seven trials, most of which had small sample sizes, so the conclusion should be viewed with caution.
Two authors disclosed receipt of fees from Amgen Inc. and Roche Inc. One author disclosed membership of the executive committee of the RED-HF (Reduction of Events with Darbepoetin Alfa in Heart Failure) trial.