Mortality and target haemoglobin concentrations in anaemic patients with chronic kidney disease treated with erythropoietin: a meta-analysis

Abstract

Background: Recombinant human erythropoietin is commonly used for treatment of anaemia. Our aim was to determine whether targeting different haemoglobin concentrations with such treatment is associated with altered all-cause mortality and cardiovascular events in patients with anaemia caused by chronic kidney disease. Methods: We did a meta-analysis of randomised controlled clinical trials that were identified in medical databases and trial registration websites. Trials were eligible for inclusion if they assessed the effects of targeting different haemoglobin concentrations in patients with anaemia caused by chronic disease who were randomly assigned to treatment with recombinant human erythropoietin, recruited at least 100 patients, and had a minimum follow-up of 12 weeks. Findings: We analysed nine randomised controlled trials that enrolled 5143 patients. There was a significantly higher risk of all-cause mortality (risk ratio 1·17, 95% CI 1·01-1·35; p=0·031) and arteriovenous access thrombosis (1·34, 1·16-1·54; p=0·0001) in the higher haemoglobin target group than in the lower haemoglobin target group in the fixed effects model without heterogeneity between studies. There was a significantly higher risk of poorly controlled blood pressure (1·27, 1·08-1·50; p=0·004) in the higher haemoglobin target group than in the lower target haemoglobin group with the fixed effects model; however, this was not significant in the random effects model (1·31, 0·97-1·78; p=0·075). The incidence of myocardial infarction was much the same in the two groups. Interpretation: To target higher haemoglobin concentrations when treating patients with anaemia caused by chronic kidney disease with recombinant human erythropoietin puts such patients at increased risk of death. Current guidelines do not include an upper limit for the target haemoglobin concentration; such an upper limit should be considered in future recommendations. © 2007 Elsevier Ltd. All rights reserved.