Hemoglobin Variability Does Not Predict Mortality in Peritoneal Dialysis Patients

Background: Hemoglobin variability in hemodialysis patients treated with erythropoiesisstimulating agents has been used to evaluate mortality and comorbidity.Different outcomes have been reported in American and European hemodialysis patients. There are, however, few studies of the effects of hemoglobinvariability in peritoneal dialysis patients.Methods: We investigated hemoglobin variability in 363 peritoneal dialysis patientsover 2 years to evaluate mortality and the association with comorbidity, peritonitis, and hospitalization. The hemoglobin of all patients selected for thestudy had been monitored for at least 6 months (April 2008 to September2008). We assessed hemoglobin variability as fluctuations from the targethemoglobin level (11–12.5 g/dL). We defined the following 6 patient groupson the basis of hemoglobin patterns: consistently low (< 11 g/dL), consistently target range (11–12.5 g/dL), consistently high (> 12.5 g/dL), low-amplitude fluctuation with low hemoglobin levels, low-amplitude fluctuation withhigh hemoglobin levels, and high amplitude fluctuation.Results: Only 2% of patients maintained a stable hemoglobin level within the targetrange and 46.8% of patients exhibited consistently low hemoglobin levels.After 2 years of observation, there was no difference in mortality as assessedby Kaplan–Meier analysis. There were also no differences in peritonitis andhospitalization between the 6 groups. However, the length of hospital staywas longer in the high amplitude fluctuation group (p = 0.008).Conclusion: Hemoglobin variability does not predict mortality in peritoneal dialysispatients.