The use of a target hemoglobin level of 13.5 g per deciliter (as compared with 11.3 g per deciliter) was associated with increased risk and no incremental improvement in the quality of life

This trial made the national papers, not because it fell into the journalist's favourite category of Breakthrough Miracle Cure, but into their second of Killer Drug Used by Doctors. An open-label trial of 1432 patients, in whom erythropoietin was used to raise haemoglobin to above 13.5 g/dl in 715 and over 11.3 in the others, led to an increase in the primary endpoint (of death, stroke, heart attack or hospitalization with heart failure) in those bumped up to the higher haemoglobins. So complete correction of anaemia appears to be harmful, and there's a poverty of evidence regarding correction to lower levels.

http://content.nejm.org/cgi/content/abstract/355/20/2085