Anaemia in dialysis patients requires treatment with frequent dose adjustments. There are two
current possible treatments for anaemia which are iron and erythropoietin stimulating agents
(ESA). Dosages of these medications are currently guided by a patient's ferritin levels and
haemoglobin, but these markers are known to be inaccurate. The current clinical protocol
therefore tends towards overuse of both agents which can be associated with toxicity, and the
reliance on these markers prevents retrospective assessment of treatment responsiveness. This
study is designed to investigate the factors which predict which agent would produce a better
response. Patients with a fall in haemoglobin will be given treatment with either iron or an
increased dose of ESA as they are currently, but allocated at random rather than by poorly
performing biochemical markers. The iron treated and ESA treated groups can then be analysed
for factors which predict response in o