A predictive algorithm for the management of anaemia in haemodialysis patients based on ESA pharmacodynamics: better results for less work.

BACKGROUND Many anaemia management algorithms recommend changes to erythropoiesis-stimulating agent (ESA) doses based on frequent measurement of haemoglobin levels in keeping with the ESA datasheets. We designed a predictive anaemia algorithm based on ESA pharmacodynamics, which we hoped would improve compliance with haemoglobin targets and reduce workload. METHODS A new algorithm was designed which predicted the 3-month steady-state haemoglobin concentration following a change in ESA dose and only recommended a change if it was outside the range 10.5-12.5 g/dL. Data were collected prospectively for 3 months prior and 15 months subsequent to implementing the algorithm. RESULTS A total of 214 prevalent dialysis patients were included in the audit. After 12 months, the haemoglobin concentration was 11.4 g/dL, near the midpoint of the target range, with a narrowing of the distribution (SD 1.46 to 1.25 g/dL, P < 0.0001). The proportion of patients with a haemoglobin level in the target range increased from 56% to 66% (P < 0.001) principally due to a reduction in the number of patients with high haemoglobin levels. There was no significant change in the ESA dose over the audit period. The number of prescription changes fell from 1/2.5 months to 1/6.1 months after 12 months (P < 0.001). CONCLUSIONS Switching prevalent haemodialysis patients to a predictive anaemia management algorithm improved compliance with haemoglobin targets, reduced the number of patients with high haemoglobin levels and reduced the number of ESA dose changes required.