Chloramine-induced haemolysis presenting as erythropoietin resistance

Background. In December 1996 we identified an outbreak of erythropoietin (rHuEpo) resistance requiring a substantial increase in rHuEpo dosage in one of our four haemodialysis (HD) units. The dialysate chloramine levels in this unit had risen from <0.1 p.p.m. in 1996 to 0.25-0.3 p.p.m. in 1997. In the other three HD units levels remained <0.1 p.p.m. Other parameters of water quality were within accepted standards. Methods. Monthly records of haemoglobin level and rHuEpo dose were available for 148 patients between January 1996 and May 1998. Seventy-two patients, with no recognized cause of rHuEpo resistance, were analysed in detail (August 1997 to April 1998). A subgroup of 15 patients was examined for evidence of haemolysis during HD (methaemoglobin and haptoglobin levels, reticulocyte counts and Heinz bodies). Larger carbon columns were installed in December 1997 to effect chloramine removal. Results. There was an increase in mean methaemoglobinaemia of 23% (P < 0.01) and a 21% fall in mean haptoglobin (P < 0.01) across HD, although no patient had a reticulocytosis and only one patient with G6PD deficiency had Heinz bodies. Following installation of larger carbon columns there was an 18.6% rise (P < 0.001) in mean haemoglobin level and a subsequent 25.0% reduction (P < 0.001) in mean rHuEpo dose. Intradialytic changes in methaemoglobin and haptoglobin were abolished. The dialysate chloramine levels fell to <0.1 p.p.m. Water company records subsequently revealed a sustained twofold increase in mains water chloramine from November 1996. Conclusions. This is the first report linking chloramine exposure and rHuEpo resistance, with only subtle signs of haemolysis. Unheralded changes in mains water constituents can directly affect dialysate water quality and clinical outcomes.